{"id":8643,"date":"2026-02-13T10:12:46","date_gmt":"2026-02-13T01:12:46","guid":{"rendered":"https:\/\/qard.is.tohoku.ac.jp\/T-Wave\/?p=8643"},"modified":"2026-02-13T10:12:46","modified_gmt":"2026-02-13T01:12:46","slug":"%e3%80%90%e5%ae%9f%e8%b7%b5%e7%b7%a8%e3%80%91%e9%87%8f%e5%ad%90%e3%82%a2%e3%83%8b%e3%83%bc%e3%83%aa%e3%83%b3%e3%82%b0%e3%82%92%e6%b4%bb%e7%94%a8%e3%81%97%e3%81%9f%e9%9d%9e%e8%b2%a0%e4%ba%8c%e5%80%a4","status":"publish","type":"post","link":"https:\/\/qard.is.tohoku.ac.jp\/T-Wave\/2026\/02\/13\/%e3%80%90%e5%ae%9f%e8%b7%b5%e7%b7%a8%e3%80%91%e9%87%8f%e5%ad%90%e3%82%a2%e3%83%8b%e3%83%bc%e3%83%aa%e3%83%b3%e3%82%b0%e3%82%92%e6%b4%bb%e7%94%a8%e3%81%97%e3%81%9f%e9%9d%9e%e8%b2%a0%e4%ba%8c%e5%80%a4\/","title":{"rendered":"\u3010\u5b9f\u8df5\u7de8\u3011\u91cf\u5b50\u30a2\u30cb\u30fc\u30ea\u30f3\u30b0\u3092\u6d3b\u7528\u3057\u305f\u975e\u8ca0\u4e8c\u5024\u884c\u5217\u56e0\u5b50\u5206\u89e3\u306b\u3088\u308b\u753b\u50cf\u5206\u985e"},"content":{"rendered":"<p><a href=\"https:\/\/colab.research.google.com\/github\/T-QARD\/t-wave\/blob\/main\/notebooks\/nbmf_imageclassification\/nbmf_imageclassification.ipynb\" target=\"_blank\" rel=\"noopener noreferrer\"><img decoding=\"async\" src=\"https:\/\/colab.research.google.com\/assets\/colab-badge.svg\" alt=\"Open in Colab\"><\/a><\/p>\n<div class=\"cell border-box-sizing text_cell rendered\">\n<div class=\"inner_cell\">\n<div class=\"text_cell_render border-box-sizing rendered_html\">\n<div id=\"ez-toc-container\" class=\"ez-toc-v2_0_82_2 counter-hierarchy ez-toc-counter ez-toc-white ez-toc-container-direction\">\n<div class=\"ez-toc-title-container\">\n<p class=\"ez-toc-title ez-toc-toggle\" style=\"cursor:pointer\">Table of Contents<\/p>\n<span class=\"ez-toc-title-toggle\"><a href=\"#\" class=\"ez-toc-pull-right ez-toc-btn ez-toc-btn-xs ez-toc-btn-default ez-toc-toggle\" aria-label=\"Toggle Table of Content\"><span class=\"ez-toc-js-icon-con\"><span class=\"\"><span class=\"eztoc-hide\" style=\"display:none;\">Toggle<\/span><span class=\"ez-toc-icon-toggle-span\"><svg style=\"fill: #999;color:#999\" xmlns=\"http:\/\/www.w3.org\/2000\/svg\" class=\"list-377408\" width=\"20px\" height=\"20px\" viewBox=\"0 0 24 24\" fill=\"none\"><path d=\"M6 6H4v2h2V6zm14 0H8v2h12V6zM4 11h2v2H4v-2zm16 0H8v2h12v-2zM4 16h2v2H4v-2zm16 0H8v2h12v-2z\" fill=\"currentColor\"><\/path><\/svg><svg style=\"fill: #999;color:#999\" class=\"arrow-unsorted-368013\" xmlns=\"http:\/\/www.w3.org\/2000\/svg\" width=\"10px\" height=\"10px\" viewBox=\"0 0 24 24\" version=\"1.2\" baseProfile=\"tiny\"><path d=\"M18.2 9.3l-6.2-6.3-6.2 6.3c-.2.2-.3.4-.3.7s.1.5.3.7c.2.2.4.3.7.3h11c.3 0 .5-.1.7-.3.2-.2.3-.5.3-.7s-.1-.5-.3-.7zM5.8 14.7l6.2 6.3 6.2-6.3c.2-.2.3-.5.3-.7s-.1-.5-.3-.7c-.2-.2-.4-.3-.7-.3h-11c-.3 0-.5.1-.7.3-.2.2-.3.5-.3.7s.1.5.3.7z\"\/><\/svg><\/span><\/span><\/span><\/a><\/span><\/div>\n<nav><ul class='ez-toc-list ez-toc-list-level-1 ' ><li class='ez-toc-page-1 ez-toc-heading-level-2'><a class=\"ez-toc-link ez-toc-heading-1\" href=\"https:\/\/qard.is.tohoku.ac.jp\/T-Wave\/2026\/02\/13\/%e3%80%90%e5%ae%9f%e8%b7%b5%e7%b7%a8%e3%80%91%e9%87%8f%e5%ad%90%e3%82%a2%e3%83%8b%e3%83%bc%e3%83%aa%e3%83%b3%e3%82%b0%e3%82%92%e6%b4%bb%e7%94%a8%e3%81%97%e3%81%9f%e9%9d%9e%e8%b2%a0%e4%ba%8c%e5%80%a4\/#%E6%A6%82%E8%A6%81\" >\u6982\u8981<\/a><\/li><li class='ez-toc-page-1 ez-toc-heading-level-2'><a class=\"ez-toc-link ez-toc-heading-2\" href=\"https:\/\/qard.is.tohoku.ac.jp\/T-Wave\/2026\/02\/13\/%e3%80%90%e5%ae%9f%e8%b7%b5%e7%b7%a8%e3%80%91%e9%87%8f%e5%ad%90%e3%82%a2%e3%83%8b%e3%83%bc%e3%83%aa%e3%83%b3%e3%82%b0%e3%82%92%e6%b4%bb%e7%94%a8%e3%81%97%e3%81%9f%e9%9d%9e%e8%b2%a0%e4%ba%8c%e5%80%a4\/#%E6%96%87%E7%8C%AE%E6%83%85%E5%A0%B1\" >\u6587\u732e\u60c5\u5831<\/a><\/li><li class='ez-toc-page-1 ez-toc-heading-level-2'><a class=\"ez-toc-link ez-toc-heading-3\" href=\"https:\/\/qard.is.tohoku.ac.jp\/T-Wave\/2026\/02\/13\/%e3%80%90%e5%ae%9f%e8%b7%b5%e7%b7%a8%e3%80%91%e9%87%8f%e5%ad%90%e3%82%a2%e3%83%8b%e3%83%bc%e3%83%aa%e3%83%b3%e3%82%b0%e3%82%92%e6%b4%bb%e7%94%a8%e3%81%97%e3%81%9f%e9%9d%9e%e8%b2%a0%e4%ba%8c%e5%80%a4\/#%E6%89%8B%E6%B3%95\" >\u624b\u6cd5<\/a><ul class='ez-toc-list-level-3' ><li class='ez-toc-heading-level-3'><a class=\"ez-toc-link ez-toc-heading-4\" href=\"https:\/\/qard.is.tohoku.ac.jp\/T-Wave\/2026\/02\/13\/%e3%80%90%e5%ae%9f%e8%b7%b5%e7%b7%a8%e3%80%91%e9%87%8f%e5%ad%90%e3%82%a2%e3%83%8b%e3%83%bc%e3%83%aa%e3%83%b3%e3%82%b0%e3%82%92%e6%b4%bb%e7%94%a8%e3%81%97%e3%81%9f%e9%9d%9e%e8%b2%a0%e4%ba%8c%e5%80%a4\/#%E9%9D%9E%E8%B2%A0%E4%BA%8C%E5%80%A4%E8%A1%8C%E5%88%97%E5%9B%A0%E5%AD%90%E5%88%86%E8%A7%A3%EF%BC%88NBMF%EF%BC%89\" >\u975e\u8ca0\u4e8c\u5024\u884c\u5217\u56e0\u5b50\u5206\u89e3\uff08NBMF\uff09<\/a><\/li><li class='ez-toc-page-1 ez-toc-heading-level-3'><a class=\"ez-toc-link ez-toc-heading-5\" href=\"https:\/\/qard.is.tohoku.ac.jp\/T-Wave\/2026\/02\/13\/%e3%80%90%e5%ae%9f%e8%b7%b5%e7%b7%a8%e3%80%91%e9%87%8f%e5%ad%90%e3%82%a2%e3%83%8b%e3%83%bc%e3%83%aa%e3%83%b3%e3%82%b0%e3%82%92%e6%b4%bb%e7%94%a8%e3%81%97%e3%81%9f%e9%9d%9e%e8%b2%a0%e4%ba%8c%e5%80%a4\/#RMSProp%E3%81%A7W%E3%82%92%E6%9B%B4%E6%96%B0%E3%81%99%E3%82%8B\" >RMSProp\u3067$W$\u3092\u66f4\u65b0\u3059\u308b<\/a><\/li><li class='ez-toc-page-1 ez-toc-heading-level-3'><a class=\"ez-toc-link ez-toc-heading-6\" href=\"https:\/\/qard.is.tohoku.ac.jp\/T-Wave\/2026\/02\/13\/%e3%80%90%e5%ae%9f%e8%b7%b5%e7%b7%a8%e3%80%91%e9%87%8f%e5%ad%90%e3%82%a2%e3%83%8b%e3%83%bc%e3%83%aa%e3%83%b3%e3%82%b0%e3%82%92%e6%b4%bb%e7%94%a8%e3%81%97%e3%81%9f%e9%9d%9e%e8%b2%a0%e4%ba%8c%e5%80%a4\/#%E9%87%8F%E5%AD%90%E3%82%A2%E3%83%8B%E3%83%BC%E3%83%AA%E3%83%B3%E3%82%B0%E3%81%A7H%E3%82%92%E6%9B%B4%E6%96%B0%E3%81%99%E3%82%8B\" >\u91cf\u5b50\u30a2\u30cb\u30fc\u30ea\u30f3\u30b0\u3067$H$\u3092\u66f4\u65b0\u3059\u308b<\/a><\/li><li class='ez-toc-page-1 ez-toc-heading-level-3'><a class=\"ez-toc-link ez-toc-heading-7\" href=\"https:\/\/qard.is.tohoku.ac.jp\/T-Wave\/2026\/02\/13\/%e3%80%90%e5%ae%9f%e8%b7%b5%e7%b7%a8%e3%80%91%e9%87%8f%e5%ad%90%e3%82%a2%e3%83%8b%e3%83%bc%e3%83%aa%e3%83%b3%e3%82%b0%e3%82%92%e6%b4%bb%e7%94%a8%e3%81%97%e3%81%9f%e9%9d%9e%e8%b2%a0%e4%ba%8c%e5%80%a4\/#%E5%85%A8%E7%B5%90%E5%90%88%E3%83%8B%E3%83%A5%E3%83%BC%E3%83%A9%E3%83%AB%E3%83%8D%E3%83%83%E3%83%88%E3%83%AF%E3%83%BC%E3%82%AF%EF%BC%88FCNN%EF%BC%89\" >\u5168\u7d50\u5408\u30cb\u30e5\u30fc\u30e9\u30eb\u30cd\u30c3\u30c8\u30ef\u30fc\u30af\uff08FCNN\uff09<\/a><\/li><\/ul><\/li><li class='ez-toc-page-1 ez-toc-heading-level-2'><a class=\"ez-toc-link ez-toc-heading-8\" href=\"https:\/\/qard.is.tohoku.ac.jp\/T-Wave\/2026\/02\/13\/%e3%80%90%e5%ae%9f%e8%b7%b5%e7%b7%a8%e3%80%91%e9%87%8f%e5%ad%90%e3%82%a2%e3%83%8b%e3%83%bc%e3%83%aa%e3%83%b3%e3%82%b0%e3%82%92%e6%b4%bb%e7%94%a8%e3%81%97%e3%81%9f%e9%9d%9e%e8%b2%a0%e4%ba%8c%e5%80%a4\/#NBMF%E3%82%92%E7%94%A8%E3%81%84%E3%81%A6%E7%94%BB%E5%83%8F%E5%88%86%E9%A1%9E%E5%95%8F%E9%A1%8C%E3%82%92%E8%A7%A3%E3%81%8F\" >NBMF\u3092\u7528\u3044\u3066\u753b\u50cf\u5206\u985e\u554f\u984c\u3092\u89e3\u304f<\/a><\/li><li class='ez-toc-page-1 ez-toc-heading-level-2'><a class=\"ez-toc-link ez-toc-heading-9\" href=\"https:\/\/qard.is.tohoku.ac.jp\/T-Wave\/2026\/02\/13\/%e3%80%90%e5%ae%9f%e8%b7%b5%e7%b7%a8%e3%80%91%e9%87%8f%e5%ad%90%e3%82%a2%e3%83%8b%e3%83%bc%e3%83%aa%e3%83%b3%e3%82%b0%e3%82%92%e6%b4%bb%e7%94%a8%e3%81%97%e3%81%9f%e9%9d%9e%e8%b2%a0%e4%ba%8c%e5%80%a4\/#%E5%AE%9F%E8%A3%85\" >\u5b9f\u88c5<\/a><ul class='ez-toc-list-level-3' ><li class='ez-toc-heading-level-3'><a class=\"ez-toc-link ez-toc-heading-10\" href=\"https:\/\/qard.is.tohoku.ac.jp\/T-Wave\/2026\/02\/13\/%e3%80%90%e5%ae%9f%e8%b7%b5%e7%b7%a8%e3%80%91%e9%87%8f%e5%ad%90%e3%82%a2%e3%83%8b%e3%83%bc%e3%83%aa%e3%83%b3%e3%82%b0%e3%82%92%e6%b4%bb%e7%94%a8%e3%81%97%e3%81%9f%e9%9d%9e%e8%b2%a0%e4%ba%8c%e5%80%a4\/#%E3%83%A9%E3%82%A4%E3%83%96%E3%83%A9%E3%83%AA%E3%82%92%E3%82%A4%E3%83%B3%E3%83%9D%E3%83%BC%E3%83%88%E3%81%99%E3%82%8B\" >\u30e9\u30a4\u30d6\u30e9\u30ea\u3092\u30a4\u30f3\u30dd\u30fc\u30c8\u3059\u308b<\/a><\/li><li class='ez-toc-page-1 ez-toc-heading-level-3'><a class=\"ez-toc-link ez-toc-heading-11\" href=\"https:\/\/qard.is.tohoku.ac.jp\/T-Wave\/2026\/02\/13\/%e3%80%90%e5%ae%9f%e8%b7%b5%e7%b7%a8%e3%80%91%e9%87%8f%e5%ad%90%e3%82%a2%e3%83%8b%e3%83%bc%e3%83%aa%e3%83%b3%e3%82%b0%e3%82%92%e6%b4%bb%e7%94%a8%e3%81%97%e3%81%9f%e9%9d%9e%e8%b2%a0%e4%ba%8c%e5%80%a4\/#%E3%83%87%E3%83%BC%E3%82%BF%E3%81%AE%E8%AA%AD%E3%81%BF%E8%BE%BC%E3%81%BF\" >\u30c7\u30fc\u30bf\u306e\u8aad\u307f\u8fbc\u307f<\/a><\/li><li class='ez-toc-page-1 ez-toc-heading-level-3'><a class=\"ez-toc-link ez-toc-heading-12\" href=\"https:\/\/qard.is.tohoku.ac.jp\/T-Wave\/2026\/02\/13\/%e3%80%90%e5%ae%9f%e8%b7%b5%e7%b7%a8%e3%80%91%e9%87%8f%e5%ad%90%e3%82%a2%e3%83%8b%e3%83%bc%e3%83%aa%e3%83%b3%e3%82%b0%e3%82%92%e6%b4%bb%e7%94%a8%e3%81%97%e3%81%9f%e9%9d%9e%e8%b2%a0%e4%ba%8c%e5%80%a4\/#RMSProp%E3%81%AB%E3%82%88%E3%82%8BW%E3%81%AE%E6%9B%B4%E6%96%B0%E5%87%A6%E7%90%86%E3%81%AE%E5%AE%9F%E8%A3%85\" >RMSProp\u306b\u3088\u308bW\u306e\u66f4\u65b0\u51e6\u7406\u306e\u5b9f\u88c5<\/a><\/li><li class='ez-toc-page-1 ez-toc-heading-level-3'><a class=\"ez-toc-link ez-toc-heading-13\" href=\"https:\/\/qard.is.tohoku.ac.jp\/T-Wave\/2026\/02\/13\/%e3%80%90%e5%ae%9f%e8%b7%b5%e7%b7%a8%e3%80%91%e9%87%8f%e5%ad%90%e3%82%a2%e3%83%8b%e3%83%bc%e3%83%aa%e3%83%b3%e3%82%b0%e3%82%92%e6%b4%bb%e7%94%a8%e3%81%97%e3%81%9f%e9%9d%9e%e8%b2%a0%e4%ba%8c%e5%80%a4\/#%E9%87%8F%E5%AD%90%E3%82%A2%E3%83%8B%E3%83%BC%E3%83%AA%E3%83%B3%E3%82%B0%E3%81%AB%E3%82%88%E3%82%8BH%E3%81%AE%E6%9B%B4%E6%96%B0%E5%87%A6%E7%90%86%E3%81%AE%E5%AE%9F%E8%A3%85\" >\u91cf\u5b50\u30a2\u30cb\u30fc\u30ea\u30f3\u30b0\u306b\u3088\u308bH\u306e\u66f4\u65b0\u51e6\u7406\u306e\u5b9f\u88c5<\/a><\/li><li class='ez-toc-page-1 ez-toc-heading-level-3'><a class=\"ez-toc-link ez-toc-heading-14\" href=\"https:\/\/qard.is.tohoku.ac.jp\/T-Wave\/2026\/02\/13\/%e3%80%90%e5%ae%9f%e8%b7%b5%e7%b7%a8%e3%80%91%e9%87%8f%e5%ad%90%e3%82%a2%e3%83%8b%e3%83%bc%e3%83%aa%e3%83%b3%e3%82%b0%e3%82%92%e6%b4%bb%e7%94%a8%e3%81%97%e3%81%9f%e9%9d%9e%e8%b2%a0%e4%ba%8c%e5%80%a4\/#NBMF%E3%81%AB%E3%82%88%E3%82%8B%E5%AD%A6%E7%BF%92%E3%81%A8%E4%BA%88%E6%B8%AC\" >NBMF\u306b\u3088\u308b\u5b66\u7fd2\u3068\u4e88\u6e2c<\/a><\/li><li class='ez-toc-page-1 ez-toc-heading-level-3'><a class=\"ez-toc-link ez-toc-heading-15\" href=\"https:\/\/qard.is.tohoku.ac.jp\/T-Wave\/2026\/02\/13\/%e3%80%90%e5%ae%9f%e8%b7%b5%e7%b7%a8%e3%80%91%e9%87%8f%e5%ad%90%e3%82%a2%e3%83%8b%e3%83%bc%e3%83%aa%e3%83%b3%e3%82%b0%e3%82%92%e6%b4%bb%e7%94%a8%e3%81%97%e3%81%9f%e9%9d%9e%e8%b2%a0%e4%ba%8c%e5%80%a4\/#FCNN%E3%81%AE%E5%AE%9F%E8%A3%85\" >FCNN\u306e\u5b9f\u88c5<\/a><\/li><li class='ez-toc-page-1 ez-toc-heading-level-3'><a class=\"ez-toc-link ez-toc-heading-16\" href=\"https:\/\/qard.is.tohoku.ac.jp\/T-Wave\/2026\/02\/13\/%e3%80%90%e5%ae%9f%e8%b7%b5%e7%b7%a8%e3%80%91%e9%87%8f%e5%ad%90%e3%82%a2%e3%83%8b%e3%83%bc%e3%83%aa%e3%83%b3%e3%82%b0%e3%82%92%e6%b4%bb%e7%94%a8%e3%81%97%e3%81%9f%e9%9d%9e%e8%b2%a0%e4%ba%8c%e5%80%a4\/#%E7%B5%90%E6%9E%9C%E3%82%92%E5%8F%AF%E8%A6%96%E5%8C%96%E3%81%99%E3%82%8B%E3%81%9F%E3%82%81%E3%81%AE%E9%96%A2%E6%95%B0\" >\u7d50\u679c\u3092\u53ef\u8996\u5316\u3059\u308b\u305f\u3081\u306e\u95a2\u6570<\/a><\/li><\/ul><\/li><li class='ez-toc-page-1 ez-toc-heading-level-2'><a class=\"ez-toc-link ez-toc-heading-17\" href=\"https:\/\/qard.is.tohoku.ac.jp\/T-Wave\/2026\/02\/13\/%e3%80%90%e5%ae%9f%e8%b7%b5%e7%b7%a8%e3%80%91%e9%87%8f%e5%ad%90%e3%82%a2%e3%83%8b%e3%83%bc%e3%83%aa%e3%83%b3%e3%82%b0%e3%82%92%e6%b4%bb%e7%94%a8%e3%81%97%e3%81%9f%e9%9d%9e%e8%b2%a0%e4%ba%8c%e5%80%a4\/#%E5%AE%9F%E9%A8%93\" >\u5b9f\u9a13<\/a><ul class='ez-toc-list-level-3' ><li class='ez-toc-heading-level-3'><a class=\"ez-toc-link ez-toc-heading-18\" href=\"https:\/\/qard.is.tohoku.ac.jp\/T-Wave\/2026\/02\/13\/%e3%80%90%e5%ae%9f%e8%b7%b5%e7%b7%a8%e3%80%91%e9%87%8f%e5%ad%90%e3%82%a2%e3%83%8b%e3%83%bc%e3%83%aa%e3%83%b3%e3%82%b0%e3%82%92%e6%b4%bb%e7%94%a8%e3%81%97%e3%81%9f%e9%9d%9e%e8%b2%a0%e4%ba%8c%e5%80%a4\/#%E5%AD%A6%E7%BF%92%E3%83%87%E3%83%BC%E3%82%BF%E6%95%B0%E3%81%AB%E5%AF%BE%E3%81%99%E3%82%8B%E7%B2%BE%E5%BA%A6%E3%81%A8%E4%BA%A4%E5%B7%AE%E3%82%A8%E3%83%B3%E3%83%88%E3%83%AD%E3%83%94%E3%83%BC%E8%AA%A4%E5%B7%AE%E3%81%A7NBMF%E3%81%A8FCNN%E3%81%AE%E6%AF%94%E8%BC%83%E3%82%92%E8%A1%8C%E3%81%86%E5%AE%9F%E9%A8%93\" >\u5b66\u7fd2\u30c7\u30fc\u30bf\u6570\u306b\u5bfe\u3059\u308b\u7cbe\u5ea6\u3068\u4ea4\u5dee\u30a8\u30f3\u30c8\u30ed\u30d4\u30fc\u8aa4\u5dee\u3067NBMF\u3068FCNN\u306e\u6bd4\u8f03\u3092\u884c\u3046\u5b9f\u9a13<\/a><\/li><li class='ez-toc-page-1 ez-toc-heading-level-3'><a class=\"ez-toc-link ez-toc-heading-19\" href=\"https:\/\/qard.is.tohoku.ac.jp\/T-Wave\/2026\/02\/13\/%e3%80%90%e5%ae%9f%e8%b7%b5%e7%b7%a8%e3%80%91%e9%87%8f%e5%ad%90%e3%82%a2%e3%83%8b%e3%83%bc%e3%83%aa%e3%83%b3%e3%82%b0%e3%82%92%e6%b4%bb%e7%94%a8%e3%81%97%e3%81%9f%e9%9d%9e%e8%b2%a0%e4%ba%8c%e5%80%a4\/#%E7%89%B9%E5%BE%B4%E6%95%B0%E3%81%AB%E5%AF%BE%E3%81%99%E3%82%8B%E7%B2%BE%E5%BA%A6%E3%81%A8%E4%BA%A4%E5%B7%AE%E3%82%A8%E3%83%B3%E3%83%88%E3%83%AD%E3%83%94%E3%83%BC%E8%AA%A4%E5%B7%AE%E3%81%A7NBMF%E3%81%A8FCNN%E3%81%AE%E6%AF%94%E8%BC%83%E3%82%92%E8%A1%8C%E3%81%86%E5%AE%9F%E9%A8%93\" >\u7279\u5fb4\u6570\u306b\u5bfe\u3059\u308b\u7cbe\u5ea6\u3068\u4ea4\u5dee\u30a8\u30f3\u30c8\u30ed\u30d4\u30fc\u8aa4\u5dee\u3067NBMF\u3068FCNN\u306e\u6bd4\u8f03\u3092\u884c\u3046\u5b9f\u9a13<\/a><\/li><li class='ez-toc-page-1 ez-toc-heading-level-3'><a class=\"ez-toc-link ez-toc-heading-20\" href=\"https:\/\/qard.is.tohoku.ac.jp\/T-Wave\/2026\/02\/13\/%e3%80%90%e5%ae%9f%e8%b7%b5%e7%b7%a8%e3%80%91%e9%87%8f%e5%ad%90%e3%82%a2%e3%83%8b%e3%83%bc%e3%83%aa%e3%83%b3%e3%82%b0%e3%82%92%e6%b4%bb%e7%94%a8%e3%81%97%e3%81%9f%e9%9d%9e%e8%b2%a0%e4%ba%8c%e5%80%a4\/#%E3%82%A8%E3%83%9D%E3%83%83%E3%82%AF%E6%95%B0%E3%81%AB%E5%AF%BE%E3%81%99%E3%82%8B%E7%B2%BE%E5%BA%A6%E3%81%A8%E4%BA%A4%E5%B7%AE%E3%82%A8%E3%83%B3%E3%83%88%E3%83%AD%E3%83%94%E3%83%BC%E8%AA%A4%E5%B7%AE%E3%81%A7NBMF%E3%81%A8FCNN%E3%81%AE%E6%AF%94%E8%BC%83%E3%82%92%E8%A1%8C%E3%81%86%E5%AE%9F%E9%A8%93\" >\u30a8\u30dd\u30c3\u30af\u6570\u306b\u5bfe\u3059\u308b\u7cbe\u5ea6\u3068\u4ea4\u5dee\u30a8\u30f3\u30c8\u30ed\u30d4\u30fc\u8aa4\u5dee\u3067NBMF\u3068FCNN\u306e\u6bd4\u8f03\u3092\u884c\u3046\u5b9f\u9a13<\/a><\/li><li class='ez-toc-page-1 ez-toc-heading-level-3'><a class=\"ez-toc-link ez-toc-heading-21\" href=\"https:\/\/qard.is.tohoku.ac.jp\/T-Wave\/2026\/02\/13\/%e3%80%90%e5%ae%9f%e8%b7%b5%e7%b7%a8%e3%80%91%e9%87%8f%e5%ad%90%e3%82%a2%e3%83%8b%e3%83%bc%e3%83%aa%e3%83%b3%e3%82%b0%e3%82%92%e6%b4%bb%e7%94%a8%e3%81%97%e3%81%9f%e9%9d%9e%e8%b2%a0%e4%ba%8c%e5%80%a4\/#W%E3%81%AE%E4%B8%AD%E8%BA%AB%E3%81%AE%E5%8F%AF%E8%A6%96%E5%8C%96\" >$W$\u306e\u4e2d\u8eab\u306e\u53ef\u8996\u5316<\/a><\/li><li class='ez-toc-page-1 ez-toc-heading-level-3'><a class=\"ez-toc-link ez-toc-heading-22\" href=\"https:\/\/qard.is.tohoku.ac.jp\/T-Wave\/2026\/02\/13\/%e3%80%90%e5%ae%9f%e8%b7%b5%e7%b7%a8%e3%80%91%e9%87%8f%e5%ad%90%e3%82%a2%e3%83%8b%e3%83%bc%e3%83%aa%e3%83%b3%e3%82%b0%e3%82%92%e6%b4%bb%e7%94%a8%e3%81%97%e3%81%9f%e9%9d%9e%e8%b2%a0%e4%ba%8c%e5%80%a4\/#W%E3%82%92%E7%94%A8%E3%81%84%E3%81%9F%E5%86%8D%E6%A7%8B%E6%88%90\" >$W$\u3092\u7528\u3044\u305f\u518d\u69cb\u6210<\/a><\/li><li class='ez-toc-page-1 ez-toc-heading-level-3'><a class=\"ez-toc-link ez-toc-heading-23\" href=\"https:\/\/qard.is.tohoku.ac.jp\/T-Wave\/2026\/02\/13\/%e3%80%90%e5%ae%9f%e8%b7%b5%e7%b7%a8%e3%80%91%e9%87%8f%e5%ad%90%e3%82%a2%e3%83%8b%e3%83%bc%e3%83%aa%e3%83%b3%e3%82%b0%e3%82%92%e6%b4%bb%e7%94%a8%e3%81%97%e3%81%9f%e9%9d%9e%e8%b2%a0%e4%ba%8c%e5%80%a4\/#%E8%AA%A4%E5%88%86%E9%A1%9E%E3%81%AB%E5%AF%BE%E3%81%99%E3%82%8B%E5%88%86%E6%9E%90\" >\u8aa4\u5206\u985e\u306b\u5bfe\u3059\u308b\u5206\u6790<\/a><\/li><\/ul><\/li><li class='ez-toc-page-1 ez-toc-heading-level-2'><a class=\"ez-toc-link ez-toc-heading-24\" href=\"https:\/\/qard.is.tohoku.ac.jp\/T-Wave\/2026\/02\/13\/%e3%80%90%e5%ae%9f%e8%b7%b5%e7%b7%a8%e3%80%91%e9%87%8f%e5%ad%90%e3%82%a2%e3%83%8b%e3%83%bc%e3%83%aa%e3%83%b3%e3%82%b0%e3%82%92%e6%b4%bb%e7%94%a8%e3%81%97%e3%81%9f%e9%9d%9e%e8%b2%a0%e4%ba%8c%e5%80%a4\/#%E3%81%BE%E3%81%A8%E3%82%81\" >\u307e\u3068\u3081<\/a><\/li><li class='ez-toc-page-1 ez-toc-heading-level-2'><a class=\"ez-toc-link ez-toc-heading-25\" href=\"https:\/\/qard.is.tohoku.ac.jp\/T-Wave\/2026\/02\/13\/%e3%80%90%e5%ae%9f%e8%b7%b5%e7%b7%a8%e3%80%91%e9%87%8f%e5%ad%90%e3%82%a2%e3%83%8b%e3%83%bc%e3%83%aa%e3%83%b3%e3%82%b0%e3%82%92%e6%b4%bb%e7%94%a8%e3%81%97%e3%81%9f%e9%9d%9e%e8%b2%a0%e4%ba%8c%e5%80%a4\/#%E3%81%82%E3%81%A8%E3%81%8C%E3%81%8D\" >\u3042\u3068\u304c\u304d<\/a><ul class='ez-toc-list-level-3' ><li class='ez-toc-heading-level-3'><a class=\"ez-toc-link ez-toc-heading-26\" href=\"https:\/\/qard.is.tohoku.ac.jp\/T-Wave\/2026\/02\/13\/%e3%80%90%e5%ae%9f%e8%b7%b5%e7%b7%a8%e3%80%91%e9%87%8f%e5%ad%90%e3%82%a2%e3%83%8b%e3%83%bc%e3%83%aa%e3%83%b3%e3%82%b0%e3%82%92%e6%b4%bb%e7%94%a8%e3%81%97%e3%81%9f%e9%9d%9e%e8%b2%a0%e4%ba%8c%e5%80%a4\/#%E6%9C%AC%E8%A8%98%E4%BA%8B%E3%81%AE%E4%BD%9C%E6%88%90%E8%80%85\" >\u672c\u8a18\u4e8b\u306e\u4f5c\u6210\u8005<\/a><\/li><\/ul><\/li><\/ul><\/nav><\/div>\n<h2><span class=\"ez-toc-section\" id=\"%E6%A6%82%E8%A6%81\"><\/span>\u6982\u8981<span class=\"ez-toc-section-end\"><\/span><\/h2>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing text_cell rendered\">\n<div class=\"inner_cell\">\n<div class=\"text_cell_render border-box-sizing rendered_html\">\n<p>\u89e3\u8aac\u8a18\u4e8b\u300c<a href=\"https:\/\/qard.is.tohoku.ac.jp\/T-Wave\/2025\/08\/28\/%e9%87%8f%e5%ad%90%e3%82%a2%e3%83%8b%e3%83%bc%e3%83%aa%e3%83%b3%e3%82%b0%e3%82%92%e6%b4%bb%e7%94%a8%e3%81%97%e3%81%9f%e9%9d%9e%e8%b2%a0%e4%ba%8c%e5%80%a4%e8%a1%8c%e5%88%97%e5%9b%a0%e5%ad%90%e5%88%86\/\">\u91cf\u5b50\u30a2\u30cb\u30fc\u30ea\u30f3\u30b0\u3092\u6d3b\u7528\u3057\u305f\u975e\u8ca0\u4e8c\u5024\u884c\u5217\u56e0\u5b50\u5206\u89e3\u306b\u3088\u308b\u753b\u50cf\u5206\u985e<\/a>\u300d\u3067\u306f\u3001\u975e\u8ca0\u4e8c\u5024\u884c\u5217\u56e0\u5b50\u5206\u89e3\u3092\u5fdc\u7528\u3057\u305f\u591a\u30af\u30e9\u30b9\u753b\u50cf\u5206\u985e\u30e2\u30c7\u30eb\u3092\u63d0\u6848\u3057, \u624b\u66f8\u304d\u6570\u5b57\u753b\u50cf\u3092\u5206\u985e\u3059\u308b\u3068\u3044\u3046\u554f\u984c\u306b\u5bfe\u3057\u3066\u63d0\u6848\u30e2\u30c7\u30eb\u3068\u5168\u7d50\u5408\u30cb\u30e5\u30fc\u30e9\u30eb\u30cd\u30c3\u30c8\u30ef\u30fc\u30af\u3067\u305d\u306e\u6027\u80fd\u3092\u6bd4\u8f03\u3059\u308b\u8ad6\u6587\u3092\u7d39\u4ecb\u3057\u307e\u3057\u305f\u3002\u672c\u8a18\u4e8b\u3067\u306f\u3001\u305d\u306e\u30a2\u30eb\u30b4\u30ea\u30ba\u30e0\u3092\u5b9f\u88c5\u3057\u3001\u5143\u8ad6\u6587\u306e\u518d\u73fe\u5b9f\u9a13\u3092\u884c\u3044\u307e\u3059\u3002<\/p>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing text_cell rendered\">\n<div class=\"inner_cell\">\n<div class=\"text_cell_render border-box-sizing rendered_html\">\n<h2><span class=\"ez-toc-section\" id=\"%E6%96%87%E7%8C%AE%E6%83%85%E5%A0%B1\"><\/span>\u6587\u732e\u60c5\u5831<span class=\"ez-toc-section-end\"><\/span><\/h2>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing text_cell rendered\">\n<div class=\"inner_cell\">\n<div class=\"text_cell_render border-box-sizing rendered_html\">\n<ul>\n<li>\u30bf\u30a4\u30c8\u30eb : Nonnegative\/Binary matrix factorization for image classification using quantum annealing<\/li>\n<li>\u8457\u8005 : Hinako Asaoka &amp; Kazue Kudo<\/li>\n<li>\u66f8\u8a8c\u60c5\u5831 :<br \/>\n<a href=\"https:\/\/doi.org\/10.1038\/s41598-023-43729-z\">https:\/\/doi.org\/10.1038\/s41598-023-43729-z<\/a><\/li>\n<\/ul>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing text_cell rendered\">\n<div class=\"inner_cell\">\n<div class=\"text_cell_render border-box-sizing rendered_html\">\n<h2><span class=\"ez-toc-section\" id=\"%E6%89%8B%E6%B3%95\"><\/span>\u624b\u6cd5<span class=\"ez-toc-section-end\"><\/span><\/h2>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing text_cell rendered\">\n<div class=\"inner_cell\">\n<div class=\"text_cell_render border-box-sizing rendered_html\">\n<p>\u5143\u8ad6\u6587\u3067\u63d0\u6848\u3055\u308c\u305f\u624b\u6cd5\u3067\u3042\u308b\u975e\u8ca0\u4e8c\u5024\u884c\u5217\u56e0\u5b50\u5206\u89e3\uff08NBMF\uff09\u3068\u5b9f\u9a13\u3067\u6bd4\u8f03\u3059\u308b\u969b\u306b\u7528\u3044\u308b\u624b\u6cd5\u3067\u3042\u308b\u5168\u7d50\u5408\u30cb\u30e5\u30fc\u30e9\u30eb\u30cd\u30c3\u30c8\u30ef\u30fc\u30af\uff08FCNN\uff09\u306b\u3064\u3044\u3066\u7c21\u5358\u306b\u632f\u308a\u8fd4\u308a\u307e\u3059\uff08\u8a73\u7d30\u306f<a href=\"https:\/\/qard.is.tohoku.ac.jp\/T-Wave\/2025\/08\/28\/%e9%87%8f%e5%ad%90%e3%82%a2%e3%83%8b%e3%83%bc%e3%83%aa%e3%83%b3%e3%82%b0%e3%82%92%e6%b4%bb%e7%94%a8%e3%81%97%e3%81%9f%e9%9d%9e%e8%b2%a0%e4%ba%8c%e5%80%a4%e8%a1%8c%e5%88%97%e5%9b%a0%e5%ad%90%e5%88%86\/\">\u89e3\u8aac\u8a18\u4e8b<\/a>\u3092\u3054\u89a7\u304f\u3060\u3055\u3044\uff09\u3002<\/p>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing text_cell rendered\">\n<div class=\"inner_cell\">\n<div class=\"text_cell_render border-box-sizing rendered_html\">\n<h3><span class=\"ez-toc-section\" id=\"%E9%9D%9E%E8%B2%A0%E4%BA%8C%E5%80%A4%E8%A1%8C%E5%88%97%E5%9B%A0%E5%AD%90%E5%88%86%E8%A7%A3%EF%BC%88NBMF%EF%BC%89\"><\/span>\u975e\u8ca0\u4e8c\u5024\u884c\u5217\u56e0\u5b50\u5206\u89e3\uff08NBMF\uff09<span class=\"ez-toc-section-end\"><\/span><\/h3>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing text_cell rendered\">\n<div class=\"inner_cell\">\n<div class=\"text_cell_render border-box-sizing rendered_html\">\n<p>\u975e\u8ca0\u4e8c\u5024\u884c\u5217\u56e0\u5b50\u5206\u89e3\uff08NBMF\uff09\u3068\u306f\u3001\u3042\u308b\u884c\u5217$V$\u3092\u57fa\u5e95\u884c\u5217$W$\u3068\u4fc2\u6570\u884c\u5217$H$\u306e\u7a4d\u306b\u5206\u89e3\u3059\u308b\u624b\u6cd5\u3067\u3059\uff08\u5f0f(1)\uff09\u3002<\/p>\n<p>$$V \\approx WH \\tag{1}$$<\/p>\n<p>\u3053\u306e\u3068\u304d\u3001\u5143\u306e\u884c\u5217$V$\u306f$n \\times m$\u884c\u5217\u3067\u3001\u57fa\u5e95\u884c\u5217$W$\u306f$n \\times k$\u3067\u6210\u5206\u304c\u975e\u8ca0\u5024\u306e\u884c\u5217\u3067\u4fc2\u6570\u884c\u5217$H$\u306f$k \\times m$\u3067\u6210\u5206\u304c\u4e8c\u5024\u306e\u884c\u5217\u3068\u3057\u307e\u3059\u3002<\/p>\n<p>NBMF\u3067\u306f\u5f0f(1)\u3092\u6e80\u305f\u3059\u3088\u3046\u306a$W$\u3068$H$\u3092\u6c42\u3081\u308b\u305f\u3081\u306b\u3001\u4ee5\u4e0b\u306e\u5f0f\u3092\u7528\u3044\u3066\u305d\u308c\u305e\u308c\u4ea4\u4e92\u306b\u66f4\u65b0\u3057\u307e\u3059\u3002<\/p>\n<p>$$W := \\arg\\min_{X \\in \\mathbb{R}_{+n \\times k} } \\bigl( \\| V &#8211; X H \\|_F + \\alpha \\| X \\|_F \\bigl) \\tag{2}$$<br \/>\n$$H := \\arg\\min_{X \\in \\{0,1\\}^{k \\times m}} \\bigl\\| V &#8211; W X \\bigl\\|_F \\tag{3}$$<\/p>\n<p>\u3053\u3053\u3067\u3001\u5f0f(2)\u3068\u5f0f(3)\u306b\u73fe\u308c\u308b$\\| \\cdot \\|_F$\u306f\u30d5\u30ed\u30d9\u30cb\u30a6\u30b9\u30ce\u30eb\u30e0\u3068\u8a00\u3044\u3001\u884c\u5217\u306e\u5168\u6210\u5206\u306e\u4e8c\u4e57\u548c\u306e\u5e73\u65b9\u6839\u3092\u8868\u3057\u307e\u3059\u3002<\/p>\n<p>\u5143\u8ad6\u6587\u3067\u306f\u3001\u5f8c\u8ff0\u3059\u308bRMSProp\u3092\u7528\u3044\u3066\u5f0f(2)\u3001\u91cf\u5b50\u30a2\u30cb\u30fc\u30ea\u30f3\u30b0\u3092\u7528\u3044\u3066\u5f0f(3)\u3092\u89e3\u3044\u3066$W$\u3068$H$\u3092\u6c42\u3081\u307e\u3059\u3002<\/p>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing text_cell rendered\">\n<div class=\"inner_cell\">\n<div class=\"text_cell_render border-box-sizing rendered_html\">\n<h3><span class=\"ez-toc-section\" id=\"RMSProp%E3%81%A7W%E3%82%92%E6%9B%B4%E6%96%B0%E3%81%99%E3%82%8B\"><\/span>RMSProp\u3067$W$\u3092\u66f4\u65b0\u3059\u308b<span class=\"ez-toc-section-end\"><\/span><\/h3>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing text_cell rendered\">\n<div class=\"inner_cell\">\n<div class=\"text_cell_render border-box-sizing rendered_html\">\n<p>\u5143\u8ad6\u6587\u3067\u63a1\u7528\u3055\u308c\u305f\u624b\u6cd5\u3067\u3042\u308bRMSProp\u3068\u306f\u52fe\u914d\u964d\u4e0b\u6cd5\u306e\u4e00\u7a2e\u3067\u3001\u52fe\u914d\u306e\u4e8c\u4e57\u306e\u79fb\u52d5\u5e73\u5747$\\boldsymbol{h}$\u306b\u5fdc\u3058\u3066\u5b66\u7fd2\u7387\u3092\u81ea\u52d5\u3067\u8abf\u6574\u3059\u308b\u624b\u6cd5\u3067\u3059\u3002\u52fe\u914d\u306e\u4e8c\u4e57\u306e\u79fb\u52d5\u5e73\u5747$\\boldsymbol{h}$\u306f\u3001\u3042\u308b\u6642\u70b9$t$\u306b\u304a\u3051\u308b\u52fe\u914d\u30d9\u30af\u30c8\u30eb\u3092$\\boldsymbol{g}$\u3001\u4ee5\u524d\u306e\u60c5\u5831\u306b\u5bfe\u3059\u308b\u91cd\u307f\u5b9a\u6570$\\beta$\u3068\u3059\u308b\u3068\u3001\u4ee5\u4e0b\u306e\u5f0f\u3067\u8868\u3055\u308c\u307e\u3059\u3002<\/p>\n<p>$$h_i^{t+1} = \\beta h_i^t + \\bigl( 1 &#8211; \\beta \\bigl) g_i^2 \\tag{4}$$<\/p>\n<p>\u307e\u305f\u3001\u5f0f(2)\u306b\u304a\u3044\u3066\u3001$W$\u306e\u3042\u308b\u7279\u5b9a\u306e\u884c\u3092\u30d9\u30af\u30c8\u30eb$\\boldsymbol{x}$\u306e\u8ee2\u7f6e\u3001$W$\u306e\u884c\u306b\u5bfe\u5fdc\u3059\u308b$V$\u306e\u884c\u3092\u30d9\u30af\u30c8\u30eb$\\boldsymbol{v}$\u306e\u8ee2\u7f6e\u3068\u3057\u3066\u8003\u3048\u3066\u3001\u640d\u5931\u95a2\u6570\u3092\u4ee5\u4e0b\u306e\u3088\u3046\u306b\u5b9a\u7fa9\u3057\u307e\u3059\u3002<\/p>\n<p>$$f_W(\\boldsymbol{x}) = \\| \\boldsymbol{v} &#8211; H ^{\\mathsf{T}} \\boldsymbol{x}\\|^2 + \\alpha \\| \\boldsymbol{x} \\|^2 \\tag{5}$$<\/p>\n<p>\u5f0f(4)\u304a\u3088\u3073\u5f0f(5)\u3092\u7528\u3044\u3066\u5909\u6570$\\boldsymbol{x}$\u306e\u66f4\u65b0\u5f0f\u3092\u4ee5\u4e0b\u306e\u901a\u308a\u3068\u3057\u307e\u3059\u3002\u305f\u3060\u3057\u3001\u5143\u8ad6\u6587\u306b\u304a\u3044\u3066\u3001\u5b9f\u9a13\u306b\u7528\u3044\u308b\u30c7\u30fc\u30bf\u30bb\u30c3\u30c8\u306fMNIST\u3067\u3001$W$\u306b\u5165\u308b\u5024\u30920\u4ee5\u4e0a1\u4ee5\u4e0b\u306e\u5b9f\u6570\u3068\u3059\u308b\u305f\u3081\u5f0f(7)\u306e\u3088\u3046\u306a\u5c04\u5f71\u3092\u8003\u3048\u3066\u3044\u307e\u3059\u3002\u307e\u305f\u3001$\\eta$\u306f\u5b66\u7fd2\u7387\u3067\u3001\u5f0f(4)\u306b\u304a\u3051\u308b\u52fe\u914d\u30d9\u30af\u30c8\u30eb$\\boldsymbol{g}$\u306f\u5f0f(5)\u306e\u640d\u5931\u95a2\u6570\u306e\u52fe\u914d$\\nabla f_W(\\boldsymbol{x})$\u3068\u3057\u307e\u3057\u305f\u3002<\/p>\n<p>$$\\boldsymbol{x}^{t+1} = P \\biggl[ \\boldsymbol{x}^t &#8211; \\eta \\frac{1}{\\sqrt{\\boldsymbol{h}^t + \\epsilon }} \\nabla f_W(\\boldsymbol{x}^t) \\biggl] \\tag{6}$$<\/p>\n<p>$$P\\left[x\\right] =<br \/>\n\\begin{cases}<br \/>\n0 \\qquad x \\leq 0 \\\\<br \/>\nx \\qquad 0 &lt; x &lt; 1 \\\\<br \/>\n1 \\qquad x \\geq 1<br \/>\n\\end{cases}<br \/>\n\\tag{7}$$<\/p>\n<p>\u4ee5\u4e0a\u306e\u3088\u3046\u306a\u3001\u5f0f(6)\u3068\u5f0f(7)\u3092\u7528\u3044\u308b\u3053\u3068\u3067RMSProp\u306b\u3088\u308b$W$\u306e\u66f4\u65b0\u304c\u53ef\u80fd\u3068\u306a\u308a\u307e\u3059\u3002<\/p>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing text_cell rendered\">\n<div class=\"inner_cell\">\n<div class=\"text_cell_render border-box-sizing rendered_html\">\n<h3><span class=\"ez-toc-section\" id=\"%E9%87%8F%E5%AD%90%E3%82%A2%E3%83%8B%E3%83%BC%E3%83%AA%E3%83%B3%E3%82%B0%E3%81%A7H%E3%82%92%E6%9B%B4%E6%96%B0%E3%81%99%E3%82%8B\"><\/span>\u91cf\u5b50\u30a2\u30cb\u30fc\u30ea\u30f3\u30b0\u3067$H$\u3092\u66f4\u65b0\u3059\u308b<span class=\"ez-toc-section-end\"><\/span><\/h3>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing text_cell rendered\">\n<div class=\"inner_cell\">\n<div class=\"text_cell_render border-box-sizing rendered_html\">\n<p>\u91cf\u5b50\u30a2\u30cb\u30fc\u30ea\u30f3\u30b0\u3068\u306f\u7d44\u307f\u5408\u308f\u305b\u6700\u9069\u5316\u554f\u984c\u3092\u89e3\u304f\u305f\u3081\u306e\u624b\u6cd5\u3067\u3059\u3002\u4eca\u56de\u306e\u554f\u984c\u306b\u304a\u3044\u3066\u3001\u5f0f(3)\u306f\u4ee5\u4e0b\u306e\u5f0f(8)\u3088\u3046\u306aQUBO\u5f62\u5f0f\u306b\u5909\u5f62\u3067\u304d\u308b\u305f\u3081\u91cf\u5b50\u30a2\u30cb\u30fc\u30ea\u30f3\u30b0\u30de\u30b7\u30f3\u3092\u7528\u3044\u3066\u89e3\u304f\u3053\u3068\u304c\u3067\u304d\u307e\u3059\u3002<\/p>\n<p>$$ QUBO \\bigl( \\boldsymbol{q} \\bigl) = \\sum_{i \\leq j} Q_{ij} q_i q_j \\tag{8} $$<\/p>\n<p>\u5f0f(3)\u3092\u5f0f(8)\u306e\u5f62\u3068\u306a\u308b\u3088\u3046\u306b\u5909\u5f62\u3059\u308b\u3068\u5f0f(9)\u306e\u76ee\u7684\u95a2\u6570\u304c\u6c42\u3081\u3089\u308c\u307e\u3059\u3002\u305f\u3060\u3057\u3001$H$\u306e\u3042\u308b\u7279\u5b9a\u306e\u5217\u3092\u30d9\u30af\u30c8\u30eb$\\boldsymbol{q}$\u3001$V$\u306e\u7279\u5b9a\u306e\u5217\u3092$\\boldsymbol{v}$\u3068\u3057\u307e\u3059\u3002<\/p>\n<p>$$f_H \\bigl( \\boldsymbol{q} \\bigl) = \\sum_i \\Bigl( \\sum_r W_{ri} \\bigl( W_{ri} &#8211; 2 v_r \\bigr) _Bigr) q_i + 2 \\sum_{i \\leq j} \\Bigl( \\sum_r W_{ri} W_{rj} \\Bigr) q_i q_j \\tag{9}$$<\/p>\n<p>\u3053\u306e\u3088\u3046\u306b\u76ee\u7684\u95a2\u6570\u3092\u8a2d\u5b9a\u3059\u308b\u3053\u3068\u3067\u3001$H$\u3092\u6c42\u3081\u308b\u3068\u3044\u3046\u554f\u984c\u3092\u91cf\u5b50\u30a2\u30cb\u30fc\u30ea\u30f3\u30b0\u3092\u7528\u3044\u3066\u89e3\u304f\u3053\u3068\u304c\u3067\u304d\u307e\u3059\u3002<\/p>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing text_cell rendered\">\n<div class=\"inner_cell\">\n<div class=\"text_cell_render border-box-sizing rendered_html\">\n<h3><span class=\"ez-toc-section\" id=\"%E5%85%A8%E7%B5%90%E5%90%88%E3%83%8B%E3%83%A5%E3%83%BC%E3%83%A9%E3%83%AB%E3%83%8D%E3%83%83%E3%83%88%E3%83%AF%E3%83%BC%E3%82%AF%EF%BC%88FCNN%EF%BC%89\"><\/span>\u5168\u7d50\u5408\u30cb\u30e5\u30fc\u30e9\u30eb\u30cd\u30c3\u30c8\u30ef\u30fc\u30af\uff08FCNN\uff09<span class=\"ez-toc-section-end\"><\/span><\/h3>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing text_cell rendered\">\n<div class=\"inner_cell\">\n<div class=\"text_cell_render border-box-sizing rendered_html\">\n<p>\u6bd4\u8f03\u5bfe\u8c61\u3068\u3059\u308bFCNN\u306f\u3001\u5165\u529b\u5c64\u30fb\u96a0\u308c\u5c64\u30fb\u51fa\u529b\u5c64\u304c\u3059\u3079\u3066\u5168\u7d50\u5408\u3055\u308c\u305f\u69cb\u9020\u306e\u30cb\u30e5\u30fc\u30e9\u30eb\u30cd\u30c3\u30c8\u30ef\u30fc\u30af\u3067\u3059\u3002\u4eca\u56de\u306e\u5b9f\u9a13\u3067\u306f\u3001\u96a0\u308c\u5c64\u306e\u6d3b\u6027\u5316\u95a2\u6570\u306bReLU\u3001\u51fa\u529b\u5c64\u306b\u30bd\u30d5\u30c8\u30de\u30c3\u30af\u30b9\u95a2\u6570\u3092\u4f7f\u7528\u3057\u307e\u3059\u3002\u307e\u305f\u3001\u904e\u5b66\u7fd2\u9632\u6b62\u306e\u305f\u3081\u96a0\u308c\u5c64\u306b\u306f\u30c9\u30ed\u30c3\u30d7\u30a2\u30a6\u30c8\uff08\u738720%\uff09\u3092\u9069\u7528\u3057\u307e\u3059\u3002\u3053\u306e\u30c9\u30ed\u30c3\u30d7\u30a2\u30a6\u30c8\u3068\u306f\u3001\u5b66\u7fd2\u306e\u5404\u30b9\u30c6\u30c3\u30d7\u306b\u304a\u3044\u3066\u30e9\u30f3\u30c0\u30e0\u306b20%\u306e\u30cb\u30e5\u30fc\u30ed\u30f3\u3092\u7121\u52b9\u5316\uff08\u51fa\u529b\u30920\u306b\uff09\u3059\u308b\u624b\u6cd5\u3067\u3001\u30cd\u30c3\u30c8\u30ef\u30fc\u30af\u304c\u7279\u5b9a\u306e\u30cb\u30e5\u30fc\u30ed\u30f3\u306b\u904e\u5ea6\u306b\u4f9d\u5b58\u3059\u308b\u3053\u3068\u3092\u9632\u3050\u52b9\u679c\u304c\u3042\u308a\u307e\u3059\u3002<\/p>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing text_cell rendered\">\n<div class=\"inner_cell\">\n<div class=\"text_cell_render border-box-sizing rendered_html\">\n<h2><span class=\"ez-toc-section\" id=\"NBMF%E3%82%92%E7%94%A8%E3%81%84%E3%81%A6%E7%94%BB%E5%83%8F%E5%88%86%E9%A1%9E%E5%95%8F%E9%A1%8C%E3%82%92%E8%A7%A3%E3%81%8F\"><\/span>NBMF\u3092\u7528\u3044\u3066\u753b\u50cf\u5206\u985e\u554f\u984c\u3092\u89e3\u304f<span class=\"ez-toc-section-end\"><\/span><\/h2>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing text_cell rendered\">\n<div class=\"inner_cell\">\n<div class=\"text_cell_render border-box-sizing rendered_html\">\n<p>\u3053\u3053\u3067\u306f\u3001NBMF\u3092\u7528\u3044\u3066\u753b\u50cf\u5206\u985e\u554f\u984c\u3092\u89e3\u304f\u65b9\u6cd5\u3092\u8aac\u660e\u3057\u307e\u3059\u3002<\/p>\n<p>\u5b9f\u9a13\u306b\u7528\u3044\u308b\u30c7\u30fc\u30bf\u30bb\u30c3\u30c8\u306fMNIST\u306e\u624b\u66f8\u304d\u6570\u5b57\u6587\u5b57\u3067\u3059\u3002\u307e\u305a\u3001\u5b66\u7fd2\u3068\u3057\u3066\u3001\u3053\u306e\u624b\u66f8\u304d\u6570\u5b57\u6587\u5b57\u753b\u50cf$m$\u679a\u3092\u5165\u529b\u3068\u3057\u3066\u884c\u5217$V$\u306b\u4e0e\u3048\u307e\u3059\u3002\u3053\u306e\u6642\u3001\u753b\u50cf1\u679a\u306e\u753b\u7d20\u6570\u306f$28 \\times 28$\u306e784\u3067\u3042\u308a\u3001\u753b\u50cf\u306e0\u304b\u30899\u306e\u30af\u30e9\u30b9\u60c5\u5831\u3092One-hot\u30d9\u30af\u30c8\u30eb\u3068\u3057\u3066\u672b\u5c3e\u306b\u8ffd\u52a0\u3059\u308b\u305f\u3081\u3001\u884c\u6570$n$\u306f$784 + 10$\u306e794\u3068\u306a\u308a\u307e\u3059\u3002<\/p>\n<p>\u3053\u306e\u5165\u529b\u884c\u5217$V$\u3092NBMF\u3092\u7528\u3044\u3066\u5206\u89e3\u3057\u3066$W$\u3068$H$\u3092\u5f97\u307e\u3059\u3002$W$\u3068$H$\u306e\u66f4\u65b0\u30921\u56de\u305a\u3064\u884c\u3046\u3053\u3068\u30921\u30a8\u30dd\u30c3\u30af\u306e\u5b66\u7fd2\u3068\u3057\u307e\u3059\u3002\u3053\u306e$W$\u306b\u5143\u306e\u5165\u529b\u753b\u50cf\u306e\u7279\u5fb4\u3068\u305d\u306e\u7279\u5fb4\u304b\u3089\u5f97\u3089\u308c\u308b\u30af\u30e9\u30b9\u306e\u60c5\u5831\u304c\u5b66\u7fd2\u3055\u308c\u3066\u3044\u308b\u305f\u3081\u3001\u3053\u308c\u3092\u4f7f\u3063\u3066\u30c6\u30b9\u30c8\u30c7\u30fc\u30bf\u3092\u5206\u985e\u3057\u307e\u3059\u3002<\/p>\n<p>$W$\u306f$n \\times k$\u884c\u5217\u3067$k$\u306f\u7279\u5fb4\u6570\u3067\u3059\u3002\u3053\u306e$W$\u306b\u306f\u7279\u5fb4\u6570$k$\u679a\u5206\u306e\u753b\u50cf\u306e\u7279\u5fb4\u3092\u8868\u3059$784 \\times k$\u884c\u5217$W_1$\u3068\u3001\u305d\u306e\u7279\u5fb4\u3092\u6301\u3064\u30af\u30e9\u30b9\u306e\u60c5\u5831\u3092\u8868\u3059$10 \\times k$\u884c\u5217$W_2$\u304c\u542b\u307e\u308c\u307e\u3059\u3002<\/p>\n<p>\u30c6\u30b9\u30c8\u30c7\u30fc\u30bf\u306e\u5206\u985e\u3067\u306f\u307e\u305a\u3001\u30c6\u30b9\u30c8\u30c7\u30fc\u30bf\u3068\u306a\u308b\u753b\u50cf$M$\u679a\u3092\u5165\u529b\u884c\u5217$V_{test}$\u3068\u3057\u3066\u3001\u5b66\u7fd2\u3055\u308c\u305f$W$\u306e$W_1$\u3092\u7528\u3044\u3066$H_{test}$\u3092\u6c42\u3081\u307e\u3059\u3002\u6b21\u306b\u3001$W_2$\u3068$H_{test}$\u3092\u4e57\u7b97\u3059\u308b\u3053\u3068\u3067\u7d50\u679c\u3068\u306a\u308b\u884c\u5217$U_{test}$\u304c\u5f97\u3089\u308c\u307e\u3059\u3002$V_{test}$\u306e\u5217\u306b\u5bfe\u5fdc\u3059\u308b$U_{test}$\u306e\u5217\u304c\u4e88\u6e2c\u3055\u308c\u305f\u30af\u30e9\u30b9\u60c5\u5831\u3067\u3042\u308a\u3001\u3053\u306e\u5217\u306b\u5bfe\u3057\u3066\u30bd\u30d5\u30c8\u30de\u30c3\u30af\u30b9\u95a2\u6570\u3092\u9069\u7528\u3057\u3066\u3001\u6700\u5927\u306e\u6210\u5206\u5024\u3092\u6301\u3064\u30a4\u30f3\u30c7\u30c3\u30af\u30b9\u3092\u4e88\u6e2c\u3055\u308c\u305f\u30af\u30e9\u30b9\u3068\u3057\u307e\u3059\u3002<\/p>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing text_cell rendered\">\n<div class=\"inner_cell\">\n<div class=\"text_cell_render border-box-sizing rendered_html\">\n<h2><span class=\"ez-toc-section\" id=\"%E5%AE%9F%E8%A3%85\"><\/span>\u5b9f\u88c5<span class=\"ez-toc-section-end\"><\/span><\/h2>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing text_cell rendered\">\n<div class=\"inner_cell\">\n<div class=\"text_cell_render border-box-sizing rendered_html\">\n<h3><span class=\"ez-toc-section\" id=\"%E3%83%A9%E3%82%A4%E3%83%96%E3%83%A9%E3%83%AA%E3%82%92%E3%82%A4%E3%83%B3%E3%83%9D%E3%83%BC%E3%83%88%E3%81%99%E3%82%8B\"><\/span>\u30e9\u30a4\u30d6\u30e9\u30ea\u3092\u30a4\u30f3\u30dd\u30fc\u30c8\u3059\u308b<span class=\"ez-toc-section-end\"><\/span><\/h3>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing text_cell rendered\">\n<div class=\"inner_cell\">\n<div class=\"text_cell_render border-box-sizing rendered_html\">\n<p>\u5b9f\u88c5\u306b\u5fc5\u8981\u306a\u30e9\u30a4\u30d6\u30e9\u30ea\u3092\u30a4\u30f3\u30dd\u30fc\u30c8\u3057\u307e\u3059\u3002<\/p>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing code_cell rendered\">\n<div class=\"input\">\n<div class=\"inner_cell\">\n<div class=\"input_area\">\n<div class=\"highlight hl-python\">\n<pre><span><\/span><span class=\"err\">!<\/span><span class=\"n\">pip<\/span> <span class=\"n\">install<\/span> <span class=\"n\">dwave<\/span><span class=\"o\">-<\/span><span class=\"n\">neal<\/span>\r\n<span class=\"err\">!<\/span><span class=\"n\">pip<\/span> <span class=\"n\">install<\/span> <span class=\"n\">openjij<\/span>\r\n<\/pre>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"output_wrapper\">\n<div class=\"output\">\n<div class=\"output_area\">\n<div class=\"prompt\"><\/div>\n<div class=\"output_subarea output_stream output_stdout output_text\">\n<pre>Requirement already satisfied: dwave-neal in \/usr\/local\/lib\/python3.12\/dist-packages (0.6.0)\r\nRequirement already satisfied: dwave-samplers&lt;2.0.0,&gt;=1.0.0 in \/usr\/local\/lib\/python3.12\/dist-packages (from dwave-neal) (1.7.0)\r\nRequirement already satisfied: numpy&lt;3.0.0,&gt;=1.21.6 in \/usr\/local\/lib\/python3.12\/dist-packages (from dwave-samplers&lt;2.0.0,&gt;=1.0.0-&gt;dwave-neal) (2.0.2)\r\nRequirement already satisfied: dimod&lt;0.13.0,&gt;=0.12.21 in \/usr\/local\/lib\/python3.12\/dist-packages (from dwave-samplers&lt;2.0.0,&gt;=1.0.0-&gt;dwave-neal) (0.12.21)\r\nRequirement already satisfied: networkx&gt;=3.0 in \/usr\/local\/lib\/python3.12\/dist-packages (from dwave-samplers&lt;2.0.0,&gt;=1.0.0-&gt;dwave-neal) (3.6.1)\r\nRequirement already satisfied: openjij in \/usr\/local\/lib\/python3.12\/dist-packages (0.11.6)\r\nRequirement already satisfied: numpy&lt;2.4.0,&gt;=1.19.3 in \/usr\/local\/lib\/python3.12\/dist-packages (from openjij) (2.0.2)\r\nRequirement already satisfied: dimod&lt;0.13.0,&gt;=0.9.11 in \/usr\/local\/lib\/python3.12\/dist-packages (from openjij) (0.12.21)\r\nRequirement already satisfied: jij-cimod&lt;1.8.0,&gt;=1.7.0 in \/usr\/local\/lib\/python3.12\/dist-packages (from openjij) (1.7.3)\r\nRequirement already satisfied: scipy&lt;1.16,&gt;=1.5.4 in \/usr\/local\/lib\/python3.12\/dist-packages (from jij-cimod&lt;1.8.0,&gt;=1.7.0-&gt;openjij) (1.15.3)\r\n<\/pre>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing code_cell rendered\">\n<div class=\"input\">\n<div class=\"inner_cell\">\n<div class=\"input_area\">\n<div class=\"highlight hl-python\">\n<pre><span><\/span><span class=\"kn\">import<\/span><span class=\"w\"> <\/span><span class=\"nn\">numpy<\/span><span class=\"w\"> <\/span><span class=\"k\">as<\/span><span class=\"w\"> <\/span><span class=\"nn\">np<\/span>\r\n<span class=\"kn\">from<\/span><span class=\"w\"> <\/span><span class=\"nn\">sklearn.datasets<\/span><span class=\"w\"> <\/span><span class=\"kn\">import<\/span> <span class=\"n\">fetch_openml<\/span>\r\n<span class=\"kn\">from<\/span><span class=\"w\"> <\/span><span class=\"nn\">sklearn.preprocessing<\/span><span class=\"w\"> <\/span><span class=\"kn\">import<\/span> <span class=\"n\">OneHotEncoder<\/span>\r\n<span class=\"kn\">from<\/span><span class=\"w\"> <\/span><span class=\"nn\">sklearn.utils<\/span><span class=\"w\"> <\/span><span class=\"kn\">import<\/span> <span class=\"n\">check_random_state<\/span>\r\n<span class=\"kn\">from<\/span><span class=\"w\"> <\/span><span class=\"nn\">sklearn.metrics<\/span><span class=\"w\"> <\/span><span class=\"kn\">import<\/span> <span class=\"n\">accuracy_score<\/span><span class=\"p\">,<\/span> <span class=\"n\">log_loss<\/span>\r\n<span class=\"kn\">import<\/span><span class=\"w\"> <\/span><span class=\"nn\">matplotlib.pyplot<\/span><span class=\"w\"> <\/span><span class=\"k\">as<\/span><span class=\"w\"> <\/span><span class=\"nn\">plt<\/span>\r\n<span class=\"kn\">from<\/span><span class=\"w\"> <\/span><span class=\"nn\">time<\/span><span class=\"w\"> <\/span><span class=\"kn\">import<\/span> <span class=\"n\">time<\/span>\r\n<span class=\"kn\">import<\/span><span class=\"w\"> <\/span><span class=\"nn\">neal<\/span>\r\n<span class=\"kn\">from<\/span><span class=\"w\"> <\/span><span class=\"nn\">dimod<\/span><span class=\"w\"> <\/span><span class=\"kn\">import<\/span> <span class=\"n\">BinaryQuadraticModel<\/span>\r\n<span class=\"kn\">import<\/span><span class=\"w\"> <\/span><span class=\"nn\">torch<\/span>\r\n<span class=\"kn\">import<\/span><span class=\"w\"> <\/span><span class=\"nn\">torch.optim<\/span><span class=\"w\"> <\/span><span class=\"k\">as<\/span><span class=\"w\"> <\/span><span class=\"nn\">optim<\/span>\r\n<span class=\"kn\">import<\/span><span class=\"w\"> <\/span><span class=\"nn\">torch.nn<\/span><span class=\"w\"> <\/span><span class=\"k\">as<\/span><span class=\"w\"> <\/span><span class=\"nn\">nn<\/span>\r\n<span class=\"kn\">import<\/span><span class=\"w\"> <\/span><span class=\"nn\">inspect<\/span>\r\n<span class=\"kn\">import<\/span><span class=\"w\"> <\/span><span class=\"nn\">openjij<\/span><span class=\"w\"> <\/span><span class=\"k\">as<\/span><span class=\"w\"> <\/span><span class=\"nn\">oj<\/span>\r\n<span class=\"kn\">from<\/span><span class=\"w\"> <\/span><span class=\"nn\">functools<\/span><span class=\"w\"> <\/span><span class=\"kn\">import<\/span> <span class=\"n\">total_ordering<\/span>\r\n<\/pre>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing text_cell rendered\">\n<div class=\"inner_cell\">\n<div class=\"text_cell_render border-box-sizing rendered_html\">\n<h3><span class=\"ez-toc-section\" id=\"%E3%83%87%E3%83%BC%E3%82%BF%E3%81%AE%E8%AA%AD%E3%81%BF%E8%BE%BC%E3%81%BF\"><\/span>\u30c7\u30fc\u30bf\u306e\u8aad\u307f\u8fbc\u307f<span class=\"ez-toc-section-end\"><\/span><\/h3>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing text_cell rendered\">\n<div class=\"inner_cell\">\n<div class=\"text_cell_render border-box-sizing rendered_html\">\n<p>\u672c\u5b9f\u9a13\u3067\u306fMNIST\u306e\u30c7\u30fc\u30bf\u30bb\u30c3\u30c8\u3092\u7528\u3044\u308b\u306e\u3067\u3001\u305d\u306e\u8aad\u307f\u8fbc\u307f\u3068\u8a13\u7df4\u30c7\u30fc\u30bf\u3068\u30c6\u30b9\u30c8\u30c7\u30fc\u30bf\u3078\u306e\u5206\u5272\u3001\u307e\u305f\u30e9\u30d9\u30eb\u3092One-hot\u30a8\u30f3\u30b3\u30fc\u30c7\u30a3\u30f3\u30b0\u3057\u3001\u753b\u50cf\u30c7\u30fc\u30bf\u3068\u7d50\u5408\u3057\u3066\u6700\u7d42\u7684\u306a\u5165\u529b\u884c\u5217\u306e\u4f5c\u6210\u3092\u884c\u3044\u307e\u3059\u3002\u3053\u306e\u3068\u304d, \u5b66\u7fd2\u30c7\u30fc\u30bf\u306f\u305d\u308c\u305e\u308c\u306e\u30af\u30e9\u30b9\u3092\u6301\u3064\u753b\u50cf\u306e\u6570\u304c\u5747\u7b49\u306b\u306a\u308b\u3088\u3046\u306b\u4f5c\u6210\u3057\u307e\u3059\u3002<\/p>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing code_cell rendered\">\n<div class=\"input\">\n<div class=\"inner_cell\">\n<div class=\"input_area\">\n<div class=\"highlight hl-python\">\n<pre><span><\/span><span class=\"k\">def<\/span><span class=\"w\"> <\/span><span class=\"nf\">load_mnist_data<\/span><span class=\"p\">(<\/span><span class=\"n\">m_train<\/span><span class=\"o\">=<\/span><span class=\"mi\">300<\/span><span class=\"p\">,<\/span> <span class=\"n\">m_test<\/span><span class=\"o\">=<\/span><span class=\"mi\">500<\/span><span class=\"p\">,<\/span> <span class=\"n\">seed<\/span><span class=\"o\">=<\/span><span class=\"mi\">0<\/span><span class=\"p\">):<\/span>\r\n<span class=\"w\">    <\/span><span class=\"sd\">\"\"\"<\/span>\r\n<span class=\"sd\">    MNIST\u30c7\u30fc\u30bf\u30bb\u30c3\u30c8\u3092\u8aad\u307f\u8fbc\u3093\u3067\u5b66\u7fd2\u30c7\u30fc\u30bf\u3068\u8a13\u7df4\u30c7\u30fc\u30bf\u306b\u5206\u5272\u3059\u308b<\/span>\r\n<span class=\"sd\">    &lt;\u5165\u529b&gt;<\/span>\r\n<span class=\"sd\">    m_train\uff1a\u5b66\u7fd2\u30c7\u30fc\u30bf\u6570<\/span>\r\n<span class=\"sd\">    m_test\uff1a\u30c6\u30b9\u30c8\u30c7\u30fc\u30bf\u6570<\/span>\r\n<span class=\"sd\">    seed\uff1a\u4e71\u6570\u30b7\u30fc\u30c9<\/span>\r\n<span class=\"sd\">    &lt;\u51fa\u529b&gt;<\/span>\r\n<span class=\"sd\">    X[train_idx]\uff1a\u5b66\u7fd2\u30c7\u30fc\u30bf\u306e\u5165\u529b\u884c\u5217<\/span>\r\n<span class=\"sd\">    y[train_idx]\uff1a\u5b66\u7fd2\u30c7\u30fc\u30bf\u306e\u6b63\u89e3\u30e9\u30d9\u30eb<\/span>\r\n<span class=\"sd\">    X[test_idx]\uff1a\u30c6\u30b9\u30c8\u30c7\u30fc\u30bf\u306e\u5165\u529b\u884c\u5217<\/span>\r\n<span class=\"sd\">    y[test_idx]\uff1a\u30c6\u30b9\u30c8\u30c7\u30fc\u30bf\u306e\u6b63\u89e3\u30e9\u30d9\u30eb<\/span>\r\n<span class=\"sd\">    \"\"\"<\/span>\r\n    <span class=\"n\">mnist<\/span> <span class=\"o\">=<\/span> <span class=\"n\">fetch_openml<\/span><span class=\"p\">(<\/span><span class=\"s1\">'mnist_784'<\/span><span class=\"p\">,<\/span> <span class=\"n\">version<\/span><span class=\"o\">=<\/span><span class=\"mi\">1<\/span><span class=\"p\">,<\/span> <span class=\"n\">as_frame<\/span><span class=\"o\">=<\/span><span class=\"kc\">False<\/span><span class=\"p\">)<\/span>\r\n    <span class=\"n\">X<\/span> <span class=\"o\">=<\/span> <span class=\"n\">mnist<\/span><span class=\"p\">[<\/span><span class=\"s1\">'data'<\/span><span class=\"p\">]<\/span><span class=\"o\">.<\/span><span class=\"n\">astype<\/span><span class=\"p\">(<\/span><span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">float32<\/span><span class=\"p\">)<\/span> <span class=\"o\">\/<\/span> <span class=\"mf\">255.0<\/span>\r\n    <span class=\"n\">y<\/span> <span class=\"o\">=<\/span> <span class=\"n\">mnist<\/span><span class=\"p\">[<\/span><span class=\"s1\">'target'<\/span><span class=\"p\">]<\/span><span class=\"o\">.<\/span><span class=\"n\">astype<\/span><span class=\"p\">(<\/span><span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">int64<\/span><span class=\"p\">)<\/span>\r\n\r\n    <span class=\"n\">rng<\/span> <span class=\"o\">=<\/span> <span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">random<\/span><span class=\"o\">.<\/span><span class=\"n\">default_rng<\/span><span class=\"p\">(<\/span><span class=\"n\">seed<\/span><span class=\"p\">)<\/span>\r\n\r\n    <span class=\"n\">n_classes<\/span> <span class=\"o\">=<\/span> <span class=\"mi\">10<\/span>\r\n    <span class=\"n\">per_class_train<\/span> <span class=\"o\">=<\/span> <span class=\"n\">m_train<\/span> <span class=\"o\">\/\/<\/span> <span class=\"n\">n_classes<\/span>\r\n\r\n    <span class=\"n\">train_idx_list<\/span> <span class=\"o\">=<\/span> <span class=\"p\">[]<\/span>\r\n\r\n    <span class=\"k\">for<\/span> <span class=\"bp\">cls<\/span> <span class=\"ow\">in<\/span> <span class=\"nb\">range<\/span><span class=\"p\">(<\/span><span class=\"n\">n_classes<\/span><span class=\"p\">):<\/span>\r\n        <span class=\"n\">cls_idx<\/span> <span class=\"o\">=<\/span> <span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">where<\/span><span class=\"p\">(<\/span><span class=\"n\">y<\/span> <span class=\"o\">==<\/span> <span class=\"bp\">cls<\/span><span class=\"p\">)[<\/span><span class=\"mi\">0<\/span><span class=\"p\">]<\/span>\r\n        <span class=\"n\">rng<\/span><span class=\"o\">.<\/span><span class=\"n\">shuffle<\/span><span class=\"p\">(<\/span><span class=\"n\">cls_idx<\/span><span class=\"p\">)<\/span>\r\n        <span class=\"n\">train_idx_list<\/span><span class=\"o\">.<\/span><span class=\"n\">append<\/span><span class=\"p\">(<\/span><span class=\"n\">cls_idx<\/span><span class=\"p\">[:<\/span><span class=\"n\">per_class_train<\/span><span class=\"p\">])<\/span>\r\n\r\n    <span class=\"n\">train_idx<\/span> <span class=\"o\">=<\/span> <span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">concatenate<\/span><span class=\"p\">(<\/span><span class=\"n\">train_idx_list<\/span><span class=\"p\">)<\/span>\r\n    <span class=\"n\">rng<\/span><span class=\"o\">.<\/span><span class=\"n\">shuffle<\/span><span class=\"p\">(<\/span><span class=\"n\">train_idx<\/span><span class=\"p\">)<\/span>\r\n\r\n    <span class=\"n\">used<\/span> <span class=\"o\">=<\/span> <span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">zeros<\/span><span class=\"p\">(<\/span><span class=\"nb\">len<\/span><span class=\"p\">(<\/span><span class=\"n\">X<\/span><span class=\"p\">),<\/span> <span class=\"n\">dtype<\/span><span class=\"o\">=<\/span><span class=\"nb\">bool<\/span><span class=\"p\">)<\/span>\r\n    <span class=\"n\">used<\/span><span class=\"p\">[<\/span><span class=\"n\">train_idx<\/span><span class=\"p\">]<\/span> <span class=\"o\">=<\/span> <span class=\"kc\">True<\/span>\r\n\r\n    <span class=\"n\">remaining_idx<\/span> <span class=\"o\">=<\/span> <span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">where<\/span><span class=\"p\">(<\/span><span class=\"o\">~<\/span><span class=\"n\">used<\/span><span class=\"p\">)[<\/span><span class=\"mi\">0<\/span><span class=\"p\">]<\/span>\r\n    <span class=\"n\">rng<\/span><span class=\"o\">.<\/span><span class=\"n\">shuffle<\/span><span class=\"p\">(<\/span><span class=\"n\">remaining_idx<\/span><span class=\"p\">)<\/span>\r\n\r\n    <span class=\"n\">test_idx<\/span> <span class=\"o\">=<\/span> <span class=\"n\">remaining_idx<\/span><span class=\"p\">[:<\/span><span class=\"n\">m_test<\/span><span class=\"p\">]<\/span>\r\n\r\n    <span class=\"k\">return<\/span> <span class=\"n\">X<\/span><span class=\"p\">[<\/span><span class=\"n\">train_idx<\/span><span class=\"p\">],<\/span> <span class=\"n\">y<\/span><span class=\"p\">[<\/span><span class=\"n\">train_idx<\/span><span class=\"p\">],<\/span> <span class=\"n\">X<\/span><span class=\"p\">[<\/span><span class=\"n\">test_idx<\/span><span class=\"p\">],<\/span> <span class=\"n\">y<\/span><span class=\"p\">[<\/span><span class=\"n\">test_idx<\/span><span class=\"p\">]<\/span>\r\n<\/pre>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing text_cell rendered\">\n<div class=\"inner_cell\">\n<div class=\"text_cell_render border-box-sizing rendered_html\">\n<h3><span class=\"ez-toc-section\" id=\"RMSProp%E3%81%AB%E3%82%88%E3%82%8BW%E3%81%AE%E6%9B%B4%E6%96%B0%E5%87%A6%E7%90%86%E3%81%AE%E5%AE%9F%E8%A3%85\"><\/span>RMSProp\u306b\u3088\u308bW\u306e\u66f4\u65b0\u51e6\u7406\u306e\u5b9f\u88c5<span class=\"ez-toc-section-end\"><\/span><\/h3>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing text_cell rendered\">\n<div class=\"inner_cell\">\n<div class=\"text_cell_render border-box-sizing rendered_html\">\n<p>\u307e\u305a\u3001RMSProp\u306b\u3088\u308bW\u306e\u66f4\u65b0\u3092\u3059\u308b\u305f\u3081\u306e\u95a2\u6570\u3092\u4f5c\u6210\u3057\u307e\u3059\u3002\u5f0f(6)\u306e\u51e6\u7406\u306e\u5b9f\u88c5\u3067\u3059\u304c\u3001\u3053\u306e\u5b9f\u88c5\u3067\u306f\u3001\u30d9\u30af\u30c8\u30eb\u3054\u3068\u306b\u66f4\u65b0\u3092\u884c\u3046\u306e\u3067\u306f\u306a\u304f\u3001\u884c\u5217\u6f14\u7b97\u3092\u7528\u3044\u3066\u4e00\u62ec\u3067\u66f4\u65b0\u51e6\u7406\u3092\u884c\u3063\u3066\u3044\u307e\u3059\u3002<\/p>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing code_cell rendered\">\n<div class=\"input\">\n<div class=\"inner_cell\">\n<div class=\"input_area\">\n<div class=\"highlight hl-python\">\n<pre><span><\/span><span class=\"k\">def<\/span><span class=\"w\"> <\/span><span class=\"nf\">update_W_rmsprop<\/span><span class=\"p\">(<\/span><span class=\"n\">V<\/span><span class=\"p\">,<\/span> <span class=\"n\">W<\/span><span class=\"p\">,<\/span> <span class=\"n\">H<\/span><span class=\"p\">,<\/span> <span class=\"n\">S<\/span><span class=\"p\">,<\/span>\r\n    <span class=\"n\">lr<\/span><span class=\"o\">=<\/span><span class=\"mf\">1e-2<\/span><span class=\"p\">,<\/span> <span class=\"n\">alpha<\/span><span class=\"o\">=<\/span><span class=\"mf\">1e-4<\/span><span class=\"p\">,<\/span> <span class=\"n\">beta<\/span><span class=\"o\">=<\/span><span class=\"mf\">0.99<\/span><span class=\"p\">,<\/span> <span class=\"n\">eps<\/span><span class=\"o\">=<\/span><span class=\"mf\">1e-7<\/span><span class=\"p\">,<\/span>\r\n    <span class=\"n\">tol<\/span><span class=\"o\">=<\/span><span class=\"mi\">1<\/span><span class=\"p\">,<\/span> <span class=\"n\">max_iter<\/span><span class=\"o\">=<\/span><span class=\"mi\">40<\/span>\r\n<span class=\"p\">):<\/span>\r\n<span class=\"w\">    <\/span><span class=\"sd\">\"\"\"<\/span>\r\n<span class=\"sd\">    RMSProp \u306b\u3088\u308b W \u306e\u66f4\u65b0\u3092\u3001\u52fe\u914d\u884c\u5217\u306e\u6700\u5927\u5024\u30ce\u30eb\u30e0\u304c\u4e00\u5b9a\u5024\u4ee5\u4e0b\u306b\u306a\u308b\u307e\u3067\u7e70\u308a\u8fd4\u3059<\/span>\r\n<span class=\"sd\">    \u524d\u56de\u306b\u6bd4\u3079\u3066\u52fe\u914d\u884c\u5217\u306e\u6700\u5927\u5024\u30ce\u30eb\u30e0\u306e\u8aa4\u5dee\u304c tol \u4ee5\u4e0b\u3082\u3057\u304f\u306f\u53cd\u5fa9\u56de\u6570\u304c max_iter \u306b\u9054\u3057\u305f\u3089\u7d42\u4e86<\/span>\r\n<span class=\"sd\">\u3000\u3000&lt;\u5165\u529b&gt;<\/span>\r\n<span class=\"sd\">    V\uff1a(n, m), \u5165\u529b\u884c\u5217<\/span>\r\n<span class=\"sd\">    W\uff1a(n, k), \u57fa\u5e95\u884c\u5217<\/span>\r\n<span class=\"sd\">    H\uff1a(k, m), \u4fc2\u6570\u884c\u5217<\/span>\r\n<span class=\"sd\">    S\uff1a(n, k), \u52fe\u914d\u306e\u4e8c\u4e57\u306e\u79fb\u52d5\u5e73\u5747<\/span>\r\n<span class=\"sd\">    lr\uff1afloat, \u5b66\u7fd2\u7387<\/span>\r\n<span class=\"sd\">    alpha\uff1afloat, \u6b63\u5247\u5316\u4fc2\u6570<\/span>\r\n<span class=\"sd\">    beta\uff1afloat, RMSProp\u306e\u6e1b\u8870\u7387<\/span>\r\n<span class=\"sd\">    eps\uff1afloat, \u30bc\u30ed\u9664\u7b97\u9632\u6b62\u9805<\/span>\r\n<span class=\"sd\">    tol\uff1afloat, \u52fe\u914d\u884c\u5217\u306e\u6700\u5927\u5024\u30ce\u30eb\u30e0\u306e\u95be\u5024<\/span>\r\n<span class=\"sd\">    max_iter\uff1aint, \u6700\u5927\u53cd\u5fa9\u56de\u6570<\/span>\r\n<span class=\"sd\">    &lt;\u51fa\u529b&gt;<\/span>\r\n<span class=\"sd\">    W\uff1a(n, k), \u66f4\u65b0\u5f8c\u306e\u57fa\u5e95\u884c\u5217<\/span>\r\n<span class=\"sd\">    S\uff1a(n, k), \u66f4\u65b0\u5f8c\u306e\u4e8c\u4e57\u5e73\u5747<\/span>\r\n<span class=\"sd\">    t+1\uff1aint, \u5b9f\u969b\u306b\u884c\u3063\u305f\u53cd\u5fa9\u56de\u6570<\/span>\r\n<span class=\"sd\">    \"\"\"<\/span>\r\n    <span class=\"n\">R<\/span> <span class=\"o\">=<\/span> <span class=\"n\">V<\/span> <span class=\"o\">-<\/span> <span class=\"n\">W<\/span> <span class=\"o\">@<\/span> <span class=\"n\">H<\/span>\r\n    <span class=\"n\">grad_prev<\/span> <span class=\"o\">=<\/span> <span class=\"o\">-<\/span><span class=\"p\">(<\/span><span class=\"n\">R<\/span> <span class=\"o\">@<\/span> <span class=\"n\">H<\/span><span class=\"o\">.<\/span><span class=\"n\">T<\/span><span class=\"p\">)<\/span><span class=\"o\">.<\/span><span class=\"n\">astype<\/span><span class=\"p\">(<\/span><span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">float32<\/span><span class=\"p\">)<\/span> <span class=\"o\">+<\/span> <span class=\"n\">alpha<\/span> <span class=\"o\">*<\/span> <span class=\"n\">W<\/span>\r\n    <span class=\"n\">grad_norm_prev<\/span> <span class=\"o\">=<\/span> <span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">linalg<\/span><span class=\"o\">.<\/span><span class=\"n\">norm<\/span><span class=\"p\">(<\/span><span class=\"n\">grad_prev<\/span><span class=\"p\">,<\/span> <span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">inf<\/span><span class=\"p\">)<\/span>\r\n\r\n    <span class=\"k\">for<\/span> <span class=\"n\">t<\/span> <span class=\"ow\">in<\/span> <span class=\"nb\">range<\/span><span class=\"p\">(<\/span><span class=\"n\">max_iter<\/span><span class=\"p\">):<\/span>\r\n        <span class=\"n\">S<\/span><span class=\"p\">[:]<\/span> <span class=\"o\">=<\/span> <span class=\"n\">beta<\/span> <span class=\"o\">*<\/span> <span class=\"n\">S<\/span> <span class=\"o\">+<\/span> <span class=\"p\">(<\/span><span class=\"mi\">1<\/span> <span class=\"o\">-<\/span> <span class=\"n\">beta<\/span><span class=\"p\">)<\/span> <span class=\"o\">*<\/span> <span class=\"p\">(<\/span><span class=\"n\">grad_prev<\/span> <span class=\"o\">*<\/span> <span class=\"n\">grad_prev<\/span><span class=\"p\">)<\/span>\r\n        <span class=\"n\">W_new<\/span> <span class=\"o\">=<\/span> <span class=\"n\">W<\/span> <span class=\"o\">-<\/span> <span class=\"n\">lr<\/span> <span class=\"o\">*<\/span> <span class=\"n\">grad_prev<\/span> <span class=\"o\">\/<\/span> <span class=\"p\">(<\/span><span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">sqrt<\/span><span class=\"p\">(<\/span><span class=\"n\">S<\/span> <span class=\"o\">+<\/span> <span class=\"n\">eps<\/span><span class=\"p\">))<\/span>\r\n        <span class=\"n\">W_new<\/span> <span class=\"o\">=<\/span> <span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">clip<\/span><span class=\"p\">(<\/span><span class=\"n\">W_new<\/span><span class=\"p\">,<\/span> <span class=\"mf\">0.0<\/span><span class=\"p\">,<\/span> <span class=\"mf\">1.0<\/span><span class=\"p\">)<\/span>\r\n\r\n        <span class=\"n\">col_norm<\/span> <span class=\"o\">=<\/span> <span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">linalg<\/span><span class=\"o\">.<\/span><span class=\"n\">norm<\/span><span class=\"p\">(<\/span><span class=\"n\">W_new<\/span><span class=\"p\">,<\/span> <span class=\"n\">axis<\/span><span class=\"o\">=<\/span><span class=\"mi\">0<\/span><span class=\"p\">,<\/span> <span class=\"n\">keepdims<\/span><span class=\"o\">=<\/span><span class=\"kc\">True<\/span><span class=\"p\">)<\/span> <span class=\"o\">+<\/span> <span class=\"mf\">1e-6<\/span>\r\n        <span class=\"n\">scale<\/span> <span class=\"o\">=<\/span> <span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">minimum<\/span><span class=\"p\">(<\/span><span class=\"mf\">1.0<\/span><span class=\"p\">,<\/span> <span class=\"mf\">1.0<\/span> <span class=\"o\">\/<\/span> <span class=\"n\">col_norm<\/span><span class=\"p\">)<\/span><span class=\"o\">.<\/span><span class=\"n\">astype<\/span><span class=\"p\">(<\/span><span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">float32<\/span><span class=\"p\">)<\/span>\r\n        <span class=\"n\">W_new<\/span> <span class=\"o\">=<\/span> <span class=\"p\">(<\/span><span class=\"n\">W_new<\/span> <span class=\"o\">*<\/span> <span class=\"n\">scale<\/span><span class=\"p\">)<\/span><span class=\"o\">.<\/span><span class=\"n\">astype<\/span><span class=\"p\">(<\/span><span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">float32<\/span><span class=\"p\">)<\/span>\r\n\r\n        <span class=\"n\">R<\/span> <span class=\"o\">=<\/span> <span class=\"n\">V<\/span> <span class=\"o\">-<\/span> <span class=\"n\">W_new<\/span> <span class=\"o\">@<\/span> <span class=\"n\">H<\/span>\r\n        <span class=\"n\">grad_new<\/span> <span class=\"o\">=<\/span> <span class=\"o\">-<\/span><span class=\"p\">(<\/span><span class=\"n\">R<\/span> <span class=\"o\">@<\/span> <span class=\"n\">H<\/span><span class=\"o\">.<\/span><span class=\"n\">T<\/span><span class=\"p\">)<\/span><span class=\"o\">.<\/span><span class=\"n\">astype<\/span><span class=\"p\">(<\/span><span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">float32<\/span><span class=\"p\">)<\/span> <span class=\"o\">+<\/span> <span class=\"n\">alpha<\/span> <span class=\"o\">*<\/span> <span class=\"n\">W_new<\/span>\r\n        <span class=\"n\">grad_norm_new<\/span> <span class=\"o\">=<\/span> <span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">linalg<\/span><span class=\"o\">.<\/span><span class=\"n\">norm<\/span><span class=\"p\">(<\/span><span class=\"n\">grad_new<\/span><span class=\"p\">,<\/span> <span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">inf<\/span><span class=\"p\">)<\/span>\r\n\r\n        <span class=\"n\">grad_diff<\/span> <span class=\"o\">=<\/span> <span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">abs<\/span><span class=\"p\">(<\/span><span class=\"n\">grad_norm_new<\/span> <span class=\"o\">-<\/span> <span class=\"n\">grad_norm_prev<\/span><span class=\"p\">)<\/span>\r\n\r\n        <span class=\"c1\"># \u8abf\u67fb\u7528\u51fa\u529b<\/span>\r\n        <span class=\"c1\"># print(f\"iter {t:4d}: grad_norm={grad_norm_new:.6f}, ratio={ratio:.6f}\")<\/span>\r\n\r\n        <span class=\"k\">if<\/span>  <span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">max<\/span><span class=\"p\">(<\/span><span class=\"n\">grad_diff<\/span><span class=\"p\">)<\/span> <span class=\"o\">&lt;<\/span> <span class=\"n\">tol<\/span><span class=\"p\">:<\/span>\r\n            <span class=\"k\">break<\/span>\r\n\r\n        <span class=\"n\">W<\/span> <span class=\"o\">=<\/span> <span class=\"n\">W_new<\/span>\r\n        <span class=\"n\">grad_prev<\/span> <span class=\"o\">=<\/span> <span class=\"n\">grad_new<\/span>\r\n        <span class=\"n\">grad_norm_prev<\/span> <span class=\"o\">=<\/span> <span class=\"n\">grad_norm_new<\/span>\r\n\r\n    <span class=\"k\">return<\/span> <span class=\"n\">W<\/span><span class=\"p\">,<\/span> <span class=\"n\">S<\/span><span class=\"p\">,<\/span> <span class=\"n\">t<\/span> <span class=\"o\">+<\/span> <span class=\"mi\">1<\/span>\r\n<\/pre>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing text_cell rendered\">\n<div class=\"inner_cell\">\n<div class=\"text_cell_render border-box-sizing rendered_html\">\n<h3><span class=\"ez-toc-section\" id=\"%E9%87%8F%E5%AD%90%E3%82%A2%E3%83%8B%E3%83%BC%E3%83%AA%E3%83%B3%E3%82%B0%E3%81%AB%E3%82%88%E3%82%8BH%E3%81%AE%E6%9B%B4%E6%96%B0%E5%87%A6%E7%90%86%E3%81%AE%E5%AE%9F%E8%A3%85\"><\/span>\u91cf\u5b50\u30a2\u30cb\u30fc\u30ea\u30f3\u30b0\u306b\u3088\u308bH\u306e\u66f4\u65b0\u51e6\u7406\u306e\u5b9f\u88c5<span class=\"ez-toc-section-end\"><\/span><\/h3>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing text_cell rendered\">\n<div class=\"inner_cell\">\n<div class=\"text_cell_render border-box-sizing rendered_html\">\n<p>\u6b21\u306b\u91cf\u5b50\u30a2\u30cb\u30fc\u30ea\u30f3\u30b0\u3092\u7528\u3044\u305fH\u306e\u66f4\u65b0\u306e\u51e6\u7406\u3092\u5b9f\u88c5\u3057\u307e\u3059\u3002\u5f0f(9)\u306e\u5b9f\u88c5\u3068\u3001\u91cf\u5b50\u30a2\u30cb\u30fc\u30ea\u30f3\u30b0\u51e6\u7406\u3092\u884c\u3044\u307e\u3059\u3002\u305f\u3060\u3057\u3001\u5b9f\u9a13\u3067\u306fD-Wave Neal\u3042\u308b\u3044\u306fOpneJij\u306eSA\u30b5\u30f3\u30d7\u30e9\u30fc\u3092\u4f7f\u7528\u3057\u307e\u3059\u3002<br \/>\n\u30b5\u30f3\u30d7\u30e9\u30fc\u3078\u306e\u5165\u529b\u306fdict\u5f62\u5f0f\u304c\u6c42\u3081\u3089\u308c\u308b\u305f\u3081\u3001\u307e\u305a\u3001\u7dda\u5f62\u9805\u3068\u4e0a\u4e09\u89d2\u884c\u5217\u304c\u4e0e\u3048\u3089\u308c\u305f\u3068\u304d\u306bQUBO\u306edict\u3068\u3057\u3066\u8fd4\u3059\u95a2\u6570\u3092\u4f5c\u6210\u3057\u307e\u3059\u3002<\/p>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing code_cell rendered\">\n<div class=\"input\">\n<div class=\"inner_cell\">\n<div class=\"input_area\">\n<div class=\"highlight hl-python\">\n<pre><span><\/span><span class=\"k\">def<\/span><span class=\"w\"> <\/span><span class=\"nf\">qubo_matrix_to_dict<\/span><span class=\"p\">(<\/span><span class=\"n\">b<\/span><span class=\"p\">,<\/span> <span class=\"n\">Q<\/span><span class=\"p\">):<\/span>\r\n<span class=\"w\">    <\/span><span class=\"sd\">\"\"\"<\/span>\r\n<span class=\"sd\">\u3000\u3000\u7dda\u5f62\u9805\u3068\u4e0a\u4e09\u89d2\u884c\u5217\u304b\u3089 QUBO \u306e\u8f9e\u66f8\u3092\u8fd4\u3059<\/span>\r\n<span class=\"sd\">    &lt;\u5165\u529b&gt;<\/span>\r\n<span class=\"sd\">    b\uff1a(k,), QUBO \u306e\u7dda\u5f62\u9805<\/span>\r\n<span class=\"sd\">    Q\uff1a(k,k), \u4e0a\u4e09\u89d2\u6210\u5206\u3092\u6301\u3064 QUBO \u884c\u5217<\/span>\r\n<span class=\"sd\">    &lt;\u51fa\u529b&gt;<\/span>\r\n<span class=\"sd\">    Q_dict\uff1a{(i,j): bias}, QUBO \u884c\u5217\u306e\u8f9e\u66f8<\/span>\r\n<span class=\"sd\">    \"\"\"<\/span>\r\n    <span class=\"n\">k<\/span> <span class=\"o\">=<\/span> <span class=\"n\">b<\/span><span class=\"o\">.<\/span><span class=\"n\">shape<\/span><span class=\"p\">[<\/span><span class=\"mi\">0<\/span><span class=\"p\">]<\/span>\r\n    <span class=\"n\">Q_dict<\/span> <span class=\"o\">=<\/span> <span class=\"p\">{}<\/span>\r\n    <span class=\"k\">for<\/span> <span class=\"n\">i<\/span> <span class=\"ow\">in<\/span> <span class=\"nb\">range<\/span><span class=\"p\">(<\/span><span class=\"n\">k<\/span><span class=\"p\">):<\/span>\r\n        <span class=\"n\">val<\/span> <span class=\"o\">=<\/span> <span class=\"nb\">float<\/span><span class=\"p\">(<\/span><span class=\"n\">b<\/span><span class=\"p\">[<\/span><span class=\"n\">i<\/span><span class=\"p\">])<\/span>\r\n        <span class=\"k\">if<\/span> <span class=\"n\">val<\/span> <span class=\"o\">!=<\/span> <span class=\"mf\">0.0<\/span><span class=\"p\">:<\/span>\r\n            <span class=\"n\">Q_dict<\/span><span class=\"p\">[(<\/span><span class=\"n\">i<\/span><span class=\"p\">,<\/span> <span class=\"n\">i<\/span><span class=\"p\">)]<\/span> <span class=\"o\">=<\/span> <span class=\"n\">val<\/span>\r\n    <span class=\"k\">for<\/span> <span class=\"n\">i<\/span> <span class=\"ow\">in<\/span> <span class=\"nb\">range<\/span><span class=\"p\">(<\/span><span class=\"n\">k<\/span><span class=\"p\">):<\/span>\r\n        <span class=\"k\">for<\/span> <span class=\"n\">j<\/span> <span class=\"ow\">in<\/span> <span class=\"nb\">range<\/span><span class=\"p\">(<\/span><span class=\"n\">i<\/span><span class=\"o\">+<\/span><span class=\"mi\">1<\/span><span class=\"p\">,<\/span> <span class=\"n\">k<\/span><span class=\"p\">):<\/span>\r\n            <span class=\"n\">val<\/span> <span class=\"o\">=<\/span> <span class=\"nb\">float<\/span><span class=\"p\">(<\/span><span class=\"n\">Q<\/span><span class=\"p\">[<\/span><span class=\"n\">i<\/span><span class=\"p\">,<\/span> <span class=\"n\">j<\/span><span class=\"p\">])<\/span>\r\n            <span class=\"k\">if<\/span> <span class=\"n\">val<\/span> <span class=\"o\">!=<\/span> <span class=\"mf\">0.0<\/span><span class=\"p\">:<\/span>\r\n                <span class=\"n\">Q_dict<\/span><span class=\"p\">[(<\/span><span class=\"n\">i<\/span><span class=\"p\">,<\/span> <span class=\"n\">j<\/span><span class=\"p\">)]<\/span> <span class=\"o\">=<\/span> <span class=\"n\">val<\/span>\r\n    <span class=\"k\">return<\/span> <span class=\"n\">Q_dict<\/span>\r\n<\/pre>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing text_cell rendered\">\n<div class=\"inner_cell\">\n<div class=\"text_cell_render border-box-sizing rendered_html\">\n<p>\u6b21\u306bQUBO\u3092\u4f5c\u6210\u3059\u308b\u305f\u3081\u306e\u8a08\u7b97\u3092\u884c\u3046\u95a2\u6570\u3092\u4f5c\u6210\u3057\u307e\u3059\u3002<\/p>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing code_cell rendered\">\n<div class=\"input\">\n<div class=\"inner_cell\">\n<div class=\"input_area\">\n<div class=\"highlight hl-python\">\n<pre><span><\/span><span class=\"k\">def<\/span><span class=\"w\"> <\/span><span class=\"nf\">precompute_qubo_terms<\/span><span class=\"p\">(<\/span><span class=\"n\">W<\/span><span class=\"p\">,<\/span> <span class=\"n\">v<\/span><span class=\"p\">,<\/span> <span class=\"n\">l1<\/span><span class=\"o\">=<\/span><span class=\"mf\">0.01<\/span><span class=\"p\">):<\/span>\r\n<span class=\"w\">    <\/span><span class=\"sd\">\"\"\"<\/span>\r\n<span class=\"sd\">    \u5f0f(9)\u306b\u57fa\u3065\u3044\u3066 QUBO \u306e\u4fc2\u6570\u3092\u8a08\u7b97\u3059\u308b<\/span>\r\n<span class=\"sd\">    f_H(q) = \u03a3_i [\u03a3_r W_ri(W_ri - 2v_r)] q_i  +  2 \u03a3_{i&lt;j} [\u03a3_r W_ri W_rj] q_i q_j<\/span>\r\n<span class=\"sd\">    \u3053\u3053\u306b\u3001\u30b9\u30d1\u30fc\u30b9\u5316\u306e\u305f\u3081\u306e\u7dda\u5f62\u9805 l1 * \u03a3 q_i \u3092\u8ffd\u52a0\u3057\u3066\u8a08\u7b97\u3092\u884c\u3046\u3002<\/span>\r\n<span class=\"sd\">    &lt;\u5165\u529b&gt;<\/span>\r\n<span class=\"sd\">    W  \uff1a(m, k), \u57fa\u5e95\u884c\u5217 (\u5f0f\u4e2d\u306e W)<\/span>\r\n<span class=\"sd\">    v  \uff1a(m,),   \u76ee\u6a19\u30d9\u30af\u30c8\u30eb (\u5f0f\u4e2d\u306e v)<\/span>\r\n<span class=\"sd\">    l1 \uff1a\u5b66\u7fd2\u7387 (\u5f0f(9)\u306b\u306f\u542b\u307e\u308c\u306a\u3044\u304c\u3001\u30b9\u30d1\u30fc\u30b9\u6027\u5236\u7d04\u3068\u3057\u3066\u7dda\u5f62\u9805\u306b\u52a0\u7b97)<\/span>\r\n<span class=\"sd\">    &lt;\u51fa\u529b&gt;<\/span>\r\n<span class=\"sd\">    b  \uff1a(k,),   \u7dda\u5f62\u9805\u306e\u4fc2\u6570 (\u5f0f(9)\u306e\u7b2c1\u9805 + l1)<\/span>\r\n<span class=\"sd\">    Q  \uff1a(k,k),  \u4e8c\u6b21\u9805\u306e\u4fc2\u6570 (\u5f0f(9)\u306e\u7b2c2\u9805, \u4e0a\u4e09\u89d2\u884c\u5217)<\/span>\r\n<span class=\"sd\">    \"\"\"<\/span>\r\n    <span class=\"c1\"># W_ri * W_rj \u306e\u548c\u3092\u8a08\u7b97 (G = W^T W)<\/span>\r\n    <span class=\"n\">G<\/span> <span class=\"o\">=<\/span> <span class=\"p\">(<\/span><span class=\"n\">W<\/span><span class=\"o\">.<\/span><span class=\"n\">T<\/span> <span class=\"o\">@<\/span> <span class=\"n\">W<\/span><span class=\"p\">)<\/span><span class=\"o\">.<\/span><span class=\"n\">astype<\/span><span class=\"p\">(<\/span><span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">float32<\/span><span class=\"p\">)<\/span>\r\n\r\n    <span class=\"c1\"># v_r * W_ri \u306e\u548c\u3092\u8a08\u7b97 (W^T v)<\/span>\r\n    <span class=\"n\">wTv<\/span> <span class=\"o\">=<\/span> <span class=\"p\">(<\/span><span class=\"n\">W<\/span><span class=\"o\">.<\/span><span class=\"n\">T<\/span> <span class=\"o\">@<\/span> <span class=\"n\">v<\/span><span class=\"p\">)<\/span><span class=\"o\">.<\/span><span class=\"n\">astype<\/span><span class=\"p\">(<\/span><span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">float32<\/span><span class=\"p\">)<\/span>\r\n\r\n    <span class=\"c1\"># --- \u7dda\u5f62\u9805 (b) \u306e\u8a08\u7b97 ---<\/span>\r\n    <span class=\"c1\"># \u5f0f(9) \u7b2c1\u9805: \u03a3_r W_ri(W_ri - 2v_r) = \u03a3 W_ri^2 - 2 \u03a3 W_ri v_r<\/span>\r\n    <span class=\"c1\"># G\u306e\u5bfe\u89d2\u6210\u5206 (np.diag(G)) \u304c \u03a3 W_ri^2 \u306b\u76f8\u5f53<\/span>\r\n    <span class=\"c1\"># \u6700\u5f8c\u306b\u30b9\u30d1\u30fc\u30b9\u5316\u9805 l1 \u3092\u52a0\u7b97<\/span>\r\n    <span class=\"n\">b<\/span> <span class=\"o\">=<\/span> <span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">diag<\/span><span class=\"p\">(<\/span><span class=\"n\">G<\/span><span class=\"p\">)<\/span> <span class=\"o\">-<\/span> <span class=\"mf\">2.0<\/span> <span class=\"o\">*<\/span> <span class=\"n\">wTv<\/span> <span class=\"o\">+<\/span> <span class=\"n\">l1<\/span>\r\n\r\n    <span class=\"c1\"># --- \u4e8c\u6b21\u9805 (Q) \u306e\u8a08\u7b97 ---<\/span>\r\n    <span class=\"c1\"># \u5f0f(9) \u7b2c2\u9805: 2 * \u03a3_{i&lt;j} (\u03a3_r W_ri W_rj) q_i q_j<\/span>\r\n    <span class=\"c1\"># G \u306e\u975e\u5bfe\u89d2\u6210\u5206 (\u4e0a\u4e09\u89d2\u90e8\u5206) \u3092\u53d6\u308a\u51fa\u3057\u3001\u4fc2\u6570 2 \u3092\u639b\u3051\u308b<\/span>\r\n    <span class=\"n\">Q<\/span> <span class=\"o\">=<\/span> <span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">triu<\/span><span class=\"p\">(<\/span><span class=\"n\">G<\/span><span class=\"p\">,<\/span> <span class=\"mi\">1<\/span><span class=\"p\">)<\/span><span class=\"o\">.<\/span><span class=\"n\">astype<\/span><span class=\"p\">(<\/span><span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">float32<\/span><span class=\"p\">)<\/span> <span class=\"o\">*<\/span> <span class=\"mf\">2.0<\/span>\r\n\r\n    <span class=\"k\">return<\/span> <span class=\"n\">b<\/span><span class=\"p\">,<\/span> <span class=\"n\">Q<\/span>\r\n<\/pre>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing text_cell rendered\">\n<div class=\"inner_cell\">\n<div class=\"text_cell_render border-box-sizing rendered_html\">\n<p>\u4f5c\u6210\u3055\u308c\u305fQUBO\u884c\u5217\u3092\u57fa\u306bSA\u30b5\u30f3\u30d7\u30e9\u30fc\u306b\u5165\u529b\u3057\u3066\u89e3\u3044\u3066\u30b3\u30b9\u30c8\u95a2\u6570\u5024\u304c\u6700\u3082\u4f4e\u304f\u306a\u308b\u3088\u3046\u306aQUBO\u3068\u305d\u306e\u6642\u306e\u30a8\u30cd\u30eb\u30ae\u30fc\u3092\u8fd4\u3059\u95a2\u6570\u3092\u4f5c\u6210\u3057\u307e\u3059\u3002<\/p>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing code_cell rendered\">\n<div class=\"input\">\n<div class=\"inner_cell\">\n<div class=\"input_area\">\n<div class=\"highlight hl-python\">\n<pre><span><\/span><span class=\"k\">def<\/span><span class=\"w\"> <\/span><span class=\"nf\">simulated_annealing_qubo<\/span><span class=\"p\">(<\/span><span class=\"n\">b<\/span><span class=\"p\">,<\/span> <span class=\"n\">Q<\/span><span class=\"p\">,<\/span> <span class=\"n\">num_reads<\/span><span class=\"o\">=<\/span><span class=\"mi\">50<\/span><span class=\"p\">,<\/span> <span class=\"n\">seed<\/span><span class=\"o\">=<\/span><span class=\"kc\">None<\/span><span class=\"p\">,<\/span> <span class=\"n\">sampler<\/span><span class=\"o\">=<\/span><span class=\"kc\">None<\/span><span class=\"p\">):<\/span>\r\n<span class=\"w\">    <\/span><span class=\"sd\">\"\"\"<\/span>\r\n<span class=\"sd\">    Simulated Annealing \u30b5\u30f3\u30d7\u30e9\u30fc\u3092\u4f7f\u3063\u3066 QUBO \u3092\u89e3\u304f<\/span>\r\n<span class=\"sd\">    &lt;\u5165\u529b&gt;<\/span>\r\n<span class=\"sd\">    b\uff1a(k,) QUBO \u306e\u7dda\u5f62\u9805<\/span>\r\n<span class=\"sd\">    Q\uff1a(k,k) \u4e0a\u4e09\u89d2 QUBO \u884c\u5217<\/span>\r\n<span class=\"sd\">    num_reads\uff1a\u30b5\u30f3\u30d7\u30e9\u30fc\u306e num_reads<\/span>\r\n<span class=\"sd\">    seed\uff1a\u4e71\u6570\u30b7\u30fc\u30c9<\/span>\r\n<span class=\"sd\">    sampler\uff1a\u65e2\u306b\u4f5c\u6210\u6e08\u307f\u306e sampler<\/span>\r\n<span class=\"sd\">    &lt;\u51fa\u529b&gt;<\/span>\r\n<span class=\"sd\">    best_q (numpy uint8 array)\uff1a\u5f97\u3089\u308c\u305f\u6700\u9069\u89e3\u30d9\u30af\u30c8\u30eb q<\/span>\r\n<span class=\"sd\">    best_energy (float)\uff1a\u6700\u9069\u89e3\u306e\u30a8\u30cd\u30eb\u30ae\u30fc<\/span>\r\n<span class=\"sd\">    elapsed_time (float)\uff1a\u30b5\u30f3\u30d7\u30ea\u30f3\u30b0\u306b\u8981\u3057\u305f\u6642\u9593\uff08\u79d2\uff09<\/span>\r\n<span class=\"sd\">    \"\"\"<\/span>\r\n    <span class=\"n\">Q_dict<\/span> <span class=\"o\">=<\/span> <span class=\"n\">qubo_matrix_to_dict<\/span><span class=\"p\">(<\/span><span class=\"n\">b<\/span><span class=\"p\">,<\/span> <span class=\"n\">Q<\/span><span class=\"p\">)<\/span>\r\n    <span class=\"n\">sampler_name<\/span> <span class=\"o\">=<\/span> <span class=\"n\">sampler<\/span><span class=\"o\">.<\/span><span class=\"vm\">__class__<\/span><span class=\"o\">.<\/span><span class=\"vm\">__module__<\/span>\r\n\r\n    <span class=\"k\">if<\/span> <span class=\"s2\">\"dwave\"<\/span> <span class=\"ow\">in<\/span> <span class=\"n\">sampler_name<\/span><span class=\"o\">.<\/span><span class=\"n\">lower<\/span><span class=\"p\">():<\/span>\r\n        <span class=\"n\">t0<\/span> <span class=\"o\">=<\/span> <span class=\"n\">time<\/span><span class=\"p\">()<\/span>\r\n        <span class=\"n\">sampleset<\/span> <span class=\"o\">=<\/span> <span class=\"n\">sampler<\/span><span class=\"o\">.<\/span><span class=\"n\">sample_qubo<\/span><span class=\"p\">(<\/span><span class=\"n\">Q_dict<\/span><span class=\"p\">)<\/span>\r\n        <span class=\"n\">t1<\/span> <span class=\"o\">=<\/span> <span class=\"n\">time<\/span><span class=\"p\">()<\/span>\r\n        <span class=\"n\">elapsed_time<\/span> <span class=\"o\">=<\/span> <span class=\"n\">t1<\/span> <span class=\"o\">-<\/span> <span class=\"n\">t0<\/span>\r\n        <span class=\"n\">first<\/span> <span class=\"o\">=<\/span> <span class=\"n\">sampleset<\/span><span class=\"o\">.<\/span><span class=\"n\">first<\/span>\r\n        <span class=\"n\">sample<\/span> <span class=\"o\">=<\/span> <span class=\"n\">first<\/span><span class=\"o\">.<\/span><span class=\"n\">sample<\/span>\r\n        <span class=\"n\">k<\/span> <span class=\"o\">=<\/span> <span class=\"n\">b<\/span><span class=\"o\">.<\/span><span class=\"n\">shape<\/span><span class=\"p\">[<\/span><span class=\"mi\">0<\/span><span class=\"p\">]<\/span>\r\n        <span class=\"n\">q_best<\/span> <span class=\"o\">=<\/span> <span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">zeros<\/span><span class=\"p\">(<\/span><span class=\"n\">k<\/span><span class=\"p\">,<\/span> <span class=\"n\">dtype<\/span><span class=\"o\">=<\/span><span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">uint8<\/span><span class=\"p\">)<\/span>\r\n        <span class=\"k\">for<\/span> <span class=\"n\">i<\/span> <span class=\"ow\">in<\/span> <span class=\"nb\">range<\/span><span class=\"p\">(<\/span><span class=\"n\">k<\/span><span class=\"p\">):<\/span>\r\n            <span class=\"n\">q_best<\/span><span class=\"p\">[<\/span><span class=\"n\">i<\/span><span class=\"p\">]<\/span> <span class=\"o\">=<\/span> <span class=\"nb\">int<\/span><span class=\"p\">(<\/span><span class=\"n\">sample<\/span><span class=\"o\">.<\/span><span class=\"n\">get<\/span><span class=\"p\">(<\/span><span class=\"n\">i<\/span><span class=\"p\">,<\/span> <span class=\"mi\">0<\/span><span class=\"p\">))<\/span>\r\n        <span class=\"n\">best_energy<\/span> <span class=\"o\">=<\/span> <span class=\"nb\">float<\/span><span class=\"p\">(<\/span><span class=\"n\">first<\/span><span class=\"o\">.<\/span><span class=\"n\">energy<\/span><span class=\"p\">)<\/span>\r\n\r\n        <span class=\"k\">return<\/span> <span class=\"n\">q_best<\/span><span class=\"p\">,<\/span> <span class=\"n\">best_energy<\/span><span class=\"p\">,<\/span> <span class=\"n\">elapsed_time<\/span>\r\n\r\n    <span class=\"k\">elif<\/span> <span class=\"s2\">\"openjij\"<\/span> <span class=\"ow\">in<\/span> <span class=\"n\">sampler_name<\/span><span class=\"o\">.<\/span><span class=\"n\">lower<\/span><span class=\"p\">():<\/span>\r\n        <span class=\"n\">best_sample<\/span> <span class=\"o\">=<\/span> <span class=\"kc\">None<\/span>\r\n        <span class=\"n\">best_energy<\/span> <span class=\"o\">=<\/span> <span class=\"nb\">float<\/span><span class=\"p\">(<\/span><span class=\"s1\">'inf'<\/span><span class=\"p\">)<\/span>\r\n        <span class=\"n\">t0<\/span> <span class=\"o\">=<\/span> <span class=\"n\">time<\/span><span class=\"p\">()<\/span>\r\n        <span class=\"n\">response<\/span> <span class=\"o\">=<\/span> <span class=\"n\">sampler<\/span><span class=\"o\">.<\/span><span class=\"n\">sample_qubo<\/span><span class=\"p\">(<\/span><span class=\"n\">Q_dict<\/span><span class=\"p\">)<\/span>\r\n        <span class=\"n\">t1<\/span> <span class=\"o\">=<\/span> <span class=\"n\">time<\/span><span class=\"p\">()<\/span>\r\n        <span class=\"n\">elapsed_time<\/span> <span class=\"o\">=<\/span> <span class=\"n\">t1<\/span> <span class=\"o\">-<\/span> <span class=\"n\">t0<\/span>\r\n        <span class=\"n\">record<\/span> <span class=\"o\">=<\/span> <span class=\"n\">response<\/span><span class=\"o\">.<\/span><span class=\"n\">record<\/span><span class=\"p\">[<\/span><span class=\"mi\">0<\/span><span class=\"p\">]<\/span>\r\n        <span class=\"n\">q<\/span> <span class=\"o\">=<\/span> <span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">array<\/span><span class=\"p\">(<\/span><span class=\"n\">record<\/span><span class=\"o\">.<\/span><span class=\"n\">sample<\/span><span class=\"p\">,<\/span> <span class=\"n\">dtype<\/span><span class=\"o\">=<\/span><span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">uint8<\/span><span class=\"p\">)<\/span>\r\n        <span class=\"n\">e<\/span> <span class=\"o\">=<\/span> <span class=\"nb\">float<\/span><span class=\"p\">(<\/span><span class=\"n\">record<\/span><span class=\"o\">.<\/span><span class=\"n\">energy<\/span><span class=\"p\">)<\/span>\r\n        <span class=\"k\">if<\/span> <span class=\"n\">e<\/span> <span class=\"o\">&lt;<\/span> <span class=\"n\">best_energy<\/span><span class=\"p\">:<\/span>\r\n            <span class=\"n\">best_energy<\/span> <span class=\"o\">=<\/span> <span class=\"n\">e<\/span>\r\n            <span class=\"n\">best_sample<\/span> <span class=\"o\">=<\/span> <span class=\"n\">q<\/span>\r\n        <span class=\"n\">q_best<\/span> <span class=\"o\">=<\/span> <span class=\"n\">best_sample<\/span>\r\n        <span class=\"k\">return<\/span> <span class=\"n\">q_best<\/span><span class=\"p\">,<\/span> <span class=\"n\">best_energy<\/span><span class=\"p\">,<\/span> <span class=\"n\">elapsed_time<\/span>\r\n    <span class=\"k\">else<\/span><span class=\"p\">:<\/span>\r\n        <span class=\"k\">raise<\/span> <span class=\"ne\">ValueError<\/span><span class=\"p\">(<\/span><span class=\"sa\">f<\/span><span class=\"s2\">\"Unknown sampler: <\/span><span class=\"si\">{<\/span><span class=\"n\">sampler_name<\/span><span class=\"si\">}<\/span><span class=\"s2\">\"<\/span><span class=\"p\">)<\/span>\r\n<\/pre>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing text_cell rendered\">\n<div class=\"inner_cell\">\n<div class=\"text_cell_render border-box-sizing rendered_html\">\n<p>\u6700\u5f8c\u306b\u3001\u4e0a\u8a18\u3067\u4f5c\u6210\u3057\u305f\u95a2\u6570\u3092\u7528\u3044\u3066H\u3092\u66f4\u65b0\u3059\u308b\u95a2\u6570\u3092\u4f5c\u6210\u3057\u307e\u3059\u3002<\/p>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing code_cell rendered\">\n<div class=\"input\">\n<div class=\"inner_cell\">\n<div class=\"input_area\">\n<div class=\"highlight hl-python\">\n<pre><span><\/span><span class=\"k\">def<\/span><span class=\"w\"> <\/span><span class=\"nf\">update_H_by_SA_qubo<\/span><span class=\"p\">(<\/span><span class=\"n\">V<\/span><span class=\"p\">,<\/span> <span class=\"n\">W<\/span><span class=\"p\">,<\/span> <span class=\"n\">H<\/span><span class=\"p\">,<\/span><span class=\"n\">l1<\/span><span class=\"o\">=<\/span><span class=\"mf\">0.01<\/span><span class=\"p\">,<\/span><span class=\"n\">rng<\/span><span class=\"o\">=<\/span><span class=\"kc\">None<\/span><span class=\"p\">,<\/span> <span class=\"n\">num_reads<\/span><span class=\"o\">=<\/span><span class=\"mi\">50<\/span><span class=\"p\">,<\/span> <span class=\"n\">sampler<\/span><span class=\"o\">=<\/span><span class=\"kc\">None<\/span><span class=\"p\">):<\/span>\r\n<span class=\"w\">    <\/span><span class=\"sd\">\"\"\"<\/span>\r\n<span class=\"sd\">    SA \u3092\u7528\u3044\u3066 H \u3092\u66f4\u65b0\u3059\u308b<\/span>\r\n<span class=\"sd\">    &lt;\u5165\u529b&gt;<\/span>\r\n<span class=\"sd\">    V\uff1a(n,m), \u76ee\u7684\u3068\u3059\u308b\u884c\u5217<\/span>\r\n<span class=\"sd\">    W\uff1a(n,k), \u57fa\u5e95\u884c\u5217<\/span>\r\n<span class=\"sd\">    H\uff1a(k,m), \u4fc2\u6570\u884c\u5217<\/span>\r\n<span class=\"sd\">    l1\uff1afloat, \u6b63\u5247\u5316\u4fc2\u6570<\/span>\r\n<span class=\"sd\">    rng\uff1a numpy.random.Generator, \u4e71\u6570\u751f\u6210\u5668<\/span>\r\n<span class=\"sd\">    num_reads\uff1aint, \u30b5\u30f3\u30d7\u30e9\u30fc\u306e num_reads<\/span>\r\n<span class=\"sd\">    sampler\uff1a\u65e2\u306b\u4f5c\u6210\u6e08\u307f\u306e sampler<\/span>\r\n<span class=\"sd\">    &lt;\u51fa\u529b&gt;<\/span>\r\n<span class=\"sd\">    H\uff1a(k,m), \u66f4\u65b0\u5f8c\u306e\u4fc2\u6570\u884c\u5217<\/span>\r\n<span class=\"sd\">    total_elapsed_time\uff1a(float), \u66f4\u65b0\u306b\u304b\u304b\u3063\u305f\u6642\u9593\uff08\u79d2\uff09<\/span>\r\n<span class=\"sd\">    \"\"\"<\/span>\r\n    <span class=\"k\">if<\/span> <span class=\"n\">rng<\/span> <span class=\"ow\">is<\/span> <span class=\"kc\">None<\/span><span class=\"p\">:<\/span>\r\n        <span class=\"n\">rng<\/span> <span class=\"o\">=<\/span> <span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">random<\/span><span class=\"o\">.<\/span><span class=\"n\">default_rng<\/span><span class=\"p\">()<\/span>\r\n    <span class=\"n\">n<\/span><span class=\"p\">,<\/span> <span class=\"n\">m<\/span> <span class=\"o\">=<\/span> <span class=\"n\">V<\/span><span class=\"o\">.<\/span><span class=\"n\">shape<\/span>\r\n    <span class=\"n\">k<\/span> <span class=\"o\">=<\/span> <span class=\"n\">W<\/span><span class=\"o\">.<\/span><span class=\"n\">shape<\/span><span class=\"p\">[<\/span><span class=\"mi\">1<\/span><span class=\"p\">]<\/span>\r\n    <span class=\"n\">total_elapsed_time<\/span> <span class=\"o\">=<\/span> <span class=\"mi\">0<\/span>\r\n\r\n    <span class=\"k\">if<\/span> <span class=\"n\">sampler<\/span> <span class=\"ow\">is<\/span> <span class=\"kc\">None<\/span><span class=\"p\">:<\/span>\r\n        <span class=\"n\">sampler<\/span> <span class=\"o\">=<\/span> <span class=\"n\">neal<\/span><span class=\"o\">.<\/span><span class=\"n\">SimulatedAnnealingSampler<\/span><span class=\"p\">()<\/span>\r\n\r\n    <span class=\"k\">for<\/span> <span class=\"n\">j<\/span> <span class=\"ow\">in<\/span> <span class=\"nb\">range<\/span><span class=\"p\">(<\/span><span class=\"n\">m<\/span><span class=\"p\">):<\/span>\r\n        <span class=\"n\">v<\/span> <span class=\"o\">=<\/span> <span class=\"n\">V<\/span><span class=\"p\">[:,<\/span> <span class=\"n\">j<\/span><span class=\"p\">]<\/span>\r\n        <span class=\"n\">b<\/span><span class=\"p\">,<\/span> <span class=\"n\">Q<\/span> <span class=\"o\">=<\/span> <span class=\"n\">precompute_qubo_terms<\/span><span class=\"p\">(<\/span><span class=\"n\">W<\/span><span class=\"p\">,<\/span> <span class=\"n\">v<\/span><span class=\"p\">,<\/span> <span class=\"n\">l1<\/span><span class=\"o\">=<\/span><span class=\"n\">l1<\/span><span class=\"p\">)<\/span>\r\n        <span class=\"n\">init_q<\/span> <span class=\"o\">=<\/span> <span class=\"n\">H<\/span><span class=\"p\">[:,<\/span> <span class=\"n\">j<\/span><span class=\"p\">]<\/span><span class=\"o\">.<\/span><span class=\"n\">astype<\/span><span class=\"p\">(<\/span><span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">uint8<\/span><span class=\"p\">)<\/span>\r\n\r\n        <span class=\"n\">seed<\/span> <span class=\"o\">=<\/span> <span class=\"nb\">int<\/span><span class=\"p\">(<\/span><span class=\"n\">rng<\/span><span class=\"o\">.<\/span><span class=\"n\">integers<\/span><span class=\"p\">(<\/span><span class=\"mi\">0<\/span><span class=\"p\">,<\/span> <span class=\"mi\">2<\/span><span class=\"o\">**<\/span><span class=\"mi\">31<\/span><span class=\"o\">-<\/span><span class=\"mi\">1<\/span><span class=\"p\">))<\/span>\r\n        <span class=\"n\">q_opt<\/span><span class=\"p\">,<\/span> <span class=\"n\">_<\/span><span class=\"p\">,<\/span> <span class=\"n\">elapsed_time<\/span> <span class=\"o\">=<\/span> <span class=\"n\">simulated_annealing_qubo<\/span><span class=\"p\">(<\/span>\r\n            <span class=\"n\">b<\/span><span class=\"o\">=<\/span><span class=\"n\">b<\/span><span class=\"p\">,<\/span> <span class=\"n\">Q<\/span><span class=\"o\">=<\/span><span class=\"n\">Q<\/span><span class=\"p\">,<\/span>\r\n            <span class=\"n\">num_reads<\/span><span class=\"o\">=<\/span><span class=\"n\">num_reads<\/span><span class=\"p\">,<\/span> <span class=\"n\">seed<\/span><span class=\"o\">=<\/span><span class=\"n\">seed<\/span><span class=\"p\">,<\/span> <span class=\"n\">sampler<\/span><span class=\"o\">=<\/span><span class=\"n\">sampler<\/span>\r\n        <span class=\"p\">)<\/span>\r\n\r\n        <span class=\"n\">H<\/span><span class=\"p\">[:,<\/span> <span class=\"n\">j<\/span><span class=\"p\">]<\/span> <span class=\"o\">=<\/span> <span class=\"n\">q_opt<\/span>\r\n        <span class=\"n\">total_elapsed_time<\/span> <span class=\"o\">+=<\/span> <span class=\"n\">elapsed_time<\/span>\r\n    <span class=\"k\">return<\/span> <span class=\"n\">H<\/span><span class=\"p\">,<\/span> <span class=\"n\">total_elapsed_time<\/span>\r\n<\/pre>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing text_cell rendered\">\n<div class=\"inner_cell\">\n<div class=\"text_cell_render border-box-sizing rendered_html\">\n<h3><span class=\"ez-toc-section\" id=\"NBMF%E3%81%AB%E3%82%88%E3%82%8B%E5%AD%A6%E7%BF%92%E3%81%A8%E4%BA%88%E6%B8%AC\"><\/span>NBMF\u306b\u3088\u308b\u5b66\u7fd2\u3068\u4e88\u6e2c<span class=\"ez-toc-section-end\"><\/span><\/h3>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing text_cell rendered\">\n<div class=\"inner_cell\">\n<div class=\"text_cell_render border-box-sizing rendered_html\">\n<p>\u307e\u305a\u3001\u5f97\u3089\u308c\u305f\u7d50\u679c\u306b\u9069\u7528\u3059\u308b\u305f\u3081\u306e\u30bd\u30d5\u30c8\u30de\u30c3\u30af\u30b9\u95a2\u6570\u306e\u4f5c\u6210\u3092\u884c\u3044\u307e\u3059\u3002<\/p>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing code_cell rendered\">\n<div class=\"input\">\n<div class=\"inner_cell\">\n<div class=\"input_area\">\n<div class=\"highlight hl-python\">\n<pre><span><\/span><span class=\"k\">def<\/span><span class=\"w\"> <\/span><span class=\"nf\">softmax<\/span><span class=\"p\">(<\/span><span class=\"n\">U<\/span><span class=\"p\">,<\/span> <span class=\"n\">axis<\/span><span class=\"o\">=<\/span><span class=\"mi\">0<\/span><span class=\"p\">):<\/span>\r\n    <span class=\"n\">axis<\/span> <span class=\"o\">=<\/span> <span class=\"mi\">0<\/span> <span class=\"k\">if<\/span> <span class=\"n\">U<\/span><span class=\"o\">.<\/span><span class=\"n\">shape<\/span><span class=\"p\">[<\/span><span class=\"mi\">0<\/span><span class=\"p\">]<\/span> <span class=\"o\">==<\/span> <span class=\"mi\">10<\/span> <span class=\"k\">else<\/span> <span class=\"mi\">1<\/span>\r\n    <span class=\"n\">U<\/span> <span class=\"o\">=<\/span> <span class=\"n\">U<\/span> <span class=\"o\">-<\/span> <span class=\"n\">U<\/span><span class=\"o\">.<\/span><span class=\"n\">max<\/span><span class=\"p\">(<\/span><span class=\"n\">axis<\/span><span class=\"o\">=<\/span><span class=\"n\">axis<\/span><span class=\"p\">,<\/span> <span class=\"n\">keepdims<\/span><span class=\"o\">=<\/span><span class=\"kc\">True<\/span><span class=\"p\">)<\/span>\r\n    <span class=\"n\">ex<\/span> <span class=\"o\">=<\/span> <span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">exp<\/span><span class=\"p\">(<\/span><span class=\"n\">U<\/span><span class=\"p\">)<\/span>\r\n    <span class=\"k\">return<\/span> <span class=\"n\">ex<\/span> <span class=\"o\">\/<\/span> <span class=\"n\">ex<\/span><span class=\"o\">.<\/span><span class=\"n\">sum<\/span><span class=\"p\">(<\/span><span class=\"n\">axis<\/span><span class=\"o\">=<\/span><span class=\"n\">axis<\/span><span class=\"p\">,<\/span> <span class=\"n\">keepdims<\/span><span class=\"o\">=<\/span><span class=\"kc\">True<\/span><span class=\"p\">)<\/span>\r\n<\/pre>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing text_cell rendered\">\n<div class=\"inner_cell\">\n<div class=\"text_cell_render border-box-sizing rendered_html\">\n<p>\u4ee5\u4e0b\u304cNBMF\u306b\u3088\u308b\u5b66\u7fd2\u30921\u56de\u5206\u884c\u3046\u305f\u3081\u306e\u30b3\u30fc\u30c9\u3067\u3059\u3002<br \/>\nNBMF\u306e\u6d41\u308c\u306f\u4ee5\u4e0b\u306b\u793a\u3059\u3068\u304a\u308a\u3067\u3059\u3002<\/p>\n<ol>\n<li>$W$\u3068$H$\u306e\u521d\u671f\u5316\uff08$W$\u306f0\u4ee5\u4e0a1\u672a\u6e80\u306e\u4e00\u69d8\u5206\u5e03\u3067$H$\u306f{0,1}\u3067\u30e9\u30f3\u30c0\u30e0\uff09<\/li>\n<li>$W$\u3092\u5c04\u5f71RMSProp\u3067\u66f4\u65b0<\/li>\n<li>$H$\u3092\u91cf\u5b50\u30a2\u30cb\u30fc\u30ea\u30f3\u30b0\u3067\u66f4\u65b0<\/li>\n<li>2\u3067\u66f4\u65b0\u3057\u305f$W$\u3092\u7528\u3044\u3066\u30c6\u30b9\u30c8\u30c7\u30fc\u30bf\u5206\u985e\u306e\u305f\u3081\u306e$H_{test}$\u4f5c\u6210<\/li>\n<li>\u5f97\u3089\u308c\u305f$H_{test}$\u3092\u7528\u3044\u3066\u30c6\u30b9\u30c8\u30c7\u30fc\u30bf\u3092\u5206\u985e\u3057\u3066\u7cbe\u5ea6\u3068\u4ea4\u5dee\u30a8\u30f3\u30c8\u30ed\u30d4\u30fc\u8aa4\u5dee\u3092\u8a55\u4fa1<\/li>\n<li>\u6307\u5b9a\u3055\u308c\u305fepochs\u6570\u5206\u7d42\u308f\u308b\u307e\u30672\u306b\u623b\u3063\u3066\u30eb\u30fc\u30d7<\/li>\n<\/ol>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing code_cell rendered\">\n<div class=\"input\">\n<div class=\"inner_cell\">\n<div class=\"input_area\">\n<div class=\"highlight hl-python\">\n<pre><span><\/span><span class=\"k\">def<\/span><span class=\"w\"> <\/span><span class=\"nf\">train_nbmf_once<\/span><span class=\"p\">(<\/span><span class=\"n\">X_train<\/span><span class=\"p\">,<\/span> <span class=\"n\">y_train<\/span><span class=\"p\">,<\/span> <span class=\"n\">X_test<\/span><span class=\"p\">,<\/span> <span class=\"n\">y_test<\/span><span class=\"p\">,<\/span>\r\n                    <span class=\"n\">k<\/span><span class=\"o\">=<\/span><span class=\"mi\">40<\/span><span class=\"p\">,<\/span> <span class=\"n\">epochs<\/span><span class=\"o\">=<\/span><span class=\"mi\">10<\/span><span class=\"p\">,<\/span> <span class=\"n\">alpha<\/span><span class=\"o\">=<\/span><span class=\"mf\">1e-4<\/span><span class=\"p\">,<\/span> <span class=\"n\">lr_W<\/span><span class=\"o\">=<\/span><span class=\"mf\">1e-2<\/span><span class=\"p\">,<\/span> <span class=\"n\">g<\/span><span class=\"o\">=<\/span><span class=\"mf\">9.0<\/span><span class=\"p\">,<\/span> <span class=\"n\">l1<\/span><span class=\"o\">=<\/span><span class=\"mf\">0.05<\/span><span class=\"p\">,<\/span> <span class=\"n\">seed<\/span><span class=\"o\">=<\/span><span class=\"mi\">0<\/span><span class=\"p\">,<\/span>\r\n                    <span class=\"n\">num_reads<\/span><span class=\"o\">=<\/span><span class=\"mi\">50<\/span><span class=\"p\">,<\/span> <span class=\"n\">verbose<\/span><span class=\"o\">=<\/span><span class=\"kc\">False<\/span><span class=\"p\">,<\/span> <span class=\"n\">sampler<\/span><span class=\"o\">=<\/span><span class=\"kc\">None<\/span><span class=\"p\">):<\/span>\r\n<span class=\"w\">    <\/span><span class=\"sd\">\"\"\"<\/span>\r\n<span class=\"sd\">    NBMF \u306e\u5b66\u7fd2\u30921\u56de\u5206\u5b9f\u884c\u3059\u308b<\/span>\r\n<span class=\"sd\">    &lt;\u5165\u529b&gt;<\/span>\r\n<span class=\"sd\">    X_train\uff1a(m_train, n_features)<\/span>\r\n<span class=\"sd\">    y_train\uff1a(m_train,)<\/span>\r\n<span class=\"sd\">    X_test\uff1a(m_test, n_features)<\/span>\r\n<span class=\"sd\">    y_test\uff1a(m_test,)<\/span>\r\n<span class=\"sd\">    k\uff1a\u7279\u5fb4\u6570<\/span>\r\n<span class=\"sd\">    alpha, lr_W\uff1aW \u306e\u6b63\u5247\u5316\u4fc2\u6570\u3068\u5b66\u7fd2\u7387<\/span>\r\n<span class=\"sd\">    g\uff1a\u30e9\u30d9\u30eb\u306b\u304b\u304b\u308b\u4fc2\u6570<\/span>\r\n<span class=\"sd\">    l1\uff1aH \u306e QUBO \u306b\u304a\u3051\u308b\u6b63\u5247\u5316\u4fc2\u6570<\/span>\r\n<span class=\"sd\">    sampler\uff1aSA \u30b5\u30f3\u30d7\u30e9\u30fc<\/span>\r\n<span class=\"sd\">    &lt;\u51fa\u529b&gt;<\/span>\r\n<span class=\"sd\">    test_acc_list\uff1a(epochs,), \u30a8\u30dd\u30c3\u30af\u3054\u3068\u306e\u7cbe\u5ea6<\/span>\r\n<span class=\"sd\">    test_ce_list\uff1a(epochs,), \u30a8\u30dd\u30c3\u30af\u3054\u3068\u306e\u4ea4\u5dee\u30a8\u30f3\u30c8\u30ed\u30d4\u30fc\u8aa4\u5dee<\/span>\r\n<span class=\"sd\">    W\uff1a(n,k), \u66f4\u65b0\u5f8c\u306e\u57fa\u5e95\u884c\u5217<\/span>\r\n<span class=\"sd\">    H\uff1a(k,m), \u66f4\u65b0\u5f8c\u306e\u4fc2\u6570\u884c\u5217<\/span>\r\n<span class=\"sd\">    \"\"\"<\/span>\r\n    <span class=\"n\">rng<\/span> <span class=\"o\">=<\/span> <span class=\"n\">check_random_state<\/span><span class=\"p\">(<\/span><span class=\"n\">seed<\/span><span class=\"p\">)<\/span>\r\n\r\n    <span class=\"c1\"># \u5b66\u7fd2\u30c7\u30fc\u30bf\u306e\u6574\u5f62<\/span>\r\n    <span class=\"n\">n_img<\/span> <span class=\"o\">=<\/span> <span class=\"n\">X_train<\/span><span class=\"o\">.<\/span><span class=\"n\">shape<\/span><span class=\"p\">[<\/span><span class=\"mi\">1<\/span><span class=\"p\">]<\/span>\r\n    <span class=\"n\">n_classes<\/span> <span class=\"o\">=<\/span> <span class=\"mi\">10<\/span>\r\n    <span class=\"n\">m_train<\/span> <span class=\"o\">=<\/span> <span class=\"n\">X_train<\/span><span class=\"o\">.<\/span><span class=\"n\">shape<\/span><span class=\"p\">[<\/span><span class=\"mi\">0<\/span><span class=\"p\">]<\/span>\r\n    <span class=\"n\">m_test<\/span> <span class=\"o\">=<\/span> <span class=\"n\">X_test<\/span><span class=\"o\">.<\/span><span class=\"n\">shape<\/span><span class=\"p\">[<\/span><span class=\"mi\">0<\/span><span class=\"p\">]<\/span>\r\n\r\n    <span class=\"c1\"># one-hot \u884c\u5217 (n_classes, m)<\/span>\r\n    <span class=\"n\">Y_train<\/span> <span class=\"o\">=<\/span> <span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">eye<\/span><span class=\"p\">(<\/span><span class=\"n\">n_classes<\/span><span class=\"p\">,<\/span> <span class=\"n\">dtype<\/span><span class=\"o\">=<\/span><span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">float32<\/span><span class=\"p\">)[<\/span><span class=\"n\">y_train<\/span><span class=\"p\">]<\/span><span class=\"o\">.<\/span><span class=\"n\">T<\/span>\r\n    <span class=\"n\">Y_test<\/span> <span class=\"o\">=<\/span> <span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">eye<\/span><span class=\"p\">(<\/span><span class=\"n\">n_classes<\/span><span class=\"p\">,<\/span> <span class=\"n\">dtype<\/span><span class=\"o\">=<\/span><span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">float32<\/span><span class=\"p\">)[<\/span><span class=\"n\">y_test<\/span><span class=\"p\">]<\/span><span class=\"o\">.<\/span><span class=\"n\">T<\/span>\r\n\r\n    <span class=\"n\">V_train<\/span> <span class=\"o\">=<\/span> <span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">vstack<\/span><span class=\"p\">([<\/span>\r\n        <span class=\"n\">X_train<\/span><span class=\"o\">.<\/span><span class=\"n\">T<\/span><span class=\"o\">.<\/span><span class=\"n\">astype<\/span><span class=\"p\">(<\/span><span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">float32<\/span><span class=\"p\">),<\/span>\r\n        <span class=\"p\">(<\/span><span class=\"n\">g<\/span> <span class=\"o\">*<\/span> <span class=\"n\">Y_train<\/span><span class=\"p\">)<\/span><span class=\"o\">.<\/span><span class=\"n\">astype<\/span><span class=\"p\">(<\/span><span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">float32<\/span><span class=\"p\">)<\/span>\r\n    <span class=\"p\">])<\/span>\r\n\r\n\r\n    <span class=\"n\">V_test_img<\/span> <span class=\"o\">=<\/span> <span class=\"n\">X_test<\/span><span class=\"o\">.<\/span><span class=\"n\">T<\/span><span class=\"o\">.<\/span><span class=\"n\">astype<\/span><span class=\"p\">(<\/span><span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">float32<\/span><span class=\"p\">)<\/span>\r\n    <span class=\"n\">n_total<\/span> <span class=\"o\">=<\/span> <span class=\"n\">V_train<\/span><span class=\"o\">.<\/span><span class=\"n\">shape<\/span><span class=\"p\">[<\/span><span class=\"mi\">0<\/span><span class=\"p\">]<\/span>\r\n\r\n    <span class=\"c1\"># \u521d\u671f\u5024<\/span>\r\n    <span class=\"n\">W<\/span> <span class=\"o\">=<\/span> <span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">clip<\/span><span class=\"p\">(<\/span><span class=\"n\">rng<\/span><span class=\"o\">.<\/span><span class=\"n\">rand<\/span><span class=\"p\">(<\/span><span class=\"n\">n_total<\/span><span class=\"p\">,<\/span> <span class=\"n\">k<\/span><span class=\"p\">)<\/span><span class=\"o\">.<\/span><span class=\"n\">astype<\/span><span class=\"p\">(<\/span><span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">float32<\/span><span class=\"p\">),<\/span> <span class=\"mf\">0.0<\/span><span class=\"p\">,<\/span> <span class=\"mf\">1.0<\/span><span class=\"p\">)<\/span>\r\n    <span class=\"n\">H<\/span> <span class=\"o\">=<\/span> <span class=\"n\">rng<\/span><span class=\"o\">.<\/span><span class=\"n\">randint<\/span><span class=\"p\">(<\/span><span class=\"mi\">0<\/span><span class=\"p\">,<\/span> <span class=\"mi\">2<\/span><span class=\"p\">,<\/span> <span class=\"n\">size<\/span><span class=\"o\">=<\/span><span class=\"p\">(<\/span><span class=\"n\">k<\/span><span class=\"p\">,<\/span> <span class=\"n\">m_train<\/span><span class=\"p\">))<\/span><span class=\"o\">.<\/span><span class=\"n\">astype<\/span><span class=\"p\">(<\/span><span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">uint8<\/span><span class=\"p\">)<\/span>\r\n\r\n    <span class=\"c1\"># RMSProp \u306b\u7528\u3044\u308b\u52fe\u914d\u306e\u521d\u671f\u5316<\/span>\r\n    <span class=\"n\">S<\/span> <span class=\"o\">=<\/span> <span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">zeros_like<\/span><span class=\"p\">(<\/span><span class=\"n\">W<\/span><span class=\"p\">,<\/span> <span class=\"n\">dtype<\/span><span class=\"o\">=<\/span><span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">float32<\/span><span class=\"p\">)<\/span>\r\n\r\n    <span class=\"n\">test_acc_list<\/span> <span class=\"o\">=<\/span> <span class=\"p\">[]<\/span>\r\n    <span class=\"n\">test_ce_list<\/span> <span class=\"o\">=<\/span> <span class=\"p\">[]<\/span>\r\n\r\n    <span class=\"n\">best_loss<\/span> <span class=\"o\">=<\/span> <span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">inf<\/span>\r\n    <span class=\"n\">no_improve<\/span> <span class=\"o\">=<\/span> <span class=\"mi\">0<\/span>\r\n    <span class=\"n\">n_iter<\/span> <span class=\"o\">=<\/span> <span class=\"mi\">0<\/span>\r\n\r\n    <span class=\"k\">for<\/span> <span class=\"n\">ep<\/span> <span class=\"ow\">in<\/span> <span class=\"nb\">range<\/span><span class=\"p\">(<\/span><span class=\"mi\">1<\/span><span class=\"p\">,<\/span> <span class=\"n\">epochs<\/span> <span class=\"o\">+<\/span> <span class=\"mi\">1<\/span><span class=\"p\">):<\/span>\r\n        <span class=\"n\">t0<\/span> <span class=\"o\">=<\/span> <span class=\"n\">time<\/span><span class=\"p\">()<\/span>\r\n\r\n        <span class=\"c1\"># W \u66f4\u65b0 (Projected RMSProp)<\/span>\r\n        <span class=\"n\">W<\/span><span class=\"p\">,<\/span> <span class=\"n\">S<\/span><span class=\"p\">,<\/span> <span class=\"n\">n_iter<\/span> <span class=\"o\">=<\/span> <span class=\"n\">update_W_rmsprop<\/span><span class=\"p\">(<\/span><span class=\"n\">V_train<\/span><span class=\"p\">,<\/span> <span class=\"n\">W<\/span><span class=\"p\">,<\/span> <span class=\"n\">H<\/span><span class=\"o\">.<\/span><span class=\"n\">astype<\/span><span class=\"p\">(<\/span><span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">float32<\/span><span class=\"p\">),<\/span> <span class=\"n\">S<\/span><span class=\"p\">,<\/span> <span class=\"n\">lr<\/span><span class=\"o\">=<\/span><span class=\"n\">lr_W<\/span><span class=\"p\">,<\/span> <span class=\"n\">alpha<\/span><span class=\"o\">=<\/span><span class=\"n\">alpha<\/span><span class=\"p\">)<\/span>\r\n\r\n        <span class=\"n\">elapsed1<\/span> <span class=\"o\">=<\/span> <span class=\"n\">time<\/span><span class=\"p\">()<\/span> <span class=\"o\">-<\/span> <span class=\"n\">t0<\/span>\r\n\r\n        <span class=\"c1\"># H \u66f4\u65b0\uff08SA\uff09<\/span>\r\n        <span class=\"n\">H<\/span><span class=\"p\">,<\/span> <span class=\"n\">H_elapsed_time<\/span> <span class=\"o\">=<\/span> <span class=\"n\">update_H_by_SA_qubo<\/span><span class=\"p\">(<\/span>\r\n            <span class=\"n\">V_train<\/span><span class=\"p\">,<\/span> <span class=\"n\">W<\/span><span class=\"p\">,<\/span> <span class=\"n\">H<\/span><span class=\"p\">,<\/span>\r\n            <span class=\"n\">l1<\/span><span class=\"o\">=<\/span><span class=\"n\">l1<\/span><span class=\"p\">,<\/span>\r\n            <span class=\"n\">rng<\/span><span class=\"o\">=<\/span><span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">random<\/span><span class=\"o\">.<\/span><span class=\"n\">default_rng<\/span><span class=\"p\">(<\/span><span class=\"n\">seed<\/span> <span class=\"o\">+<\/span> <span class=\"n\">ep<\/span><span class=\"p\">),<\/span>\r\n            <span class=\"n\">num_reads<\/span><span class=\"o\">=<\/span><span class=\"n\">num_reads<\/span><span class=\"p\">,<\/span>\r\n            <span class=\"n\">sampler<\/span><span class=\"o\">=<\/span><span class=\"n\">sampler<\/span>\r\n        <span class=\"p\">)<\/span>\r\n\r\n        <span class=\"c1\"># 3) \u30c6\u30b9\u30c8\u3067\u8a55\u4fa1\uff08\u3053\u306e\u6bb5\u968e\u306e W \u3092\u4f7f\u3063\u3066 SA \u306b\u3088\u308a H_test \u3092\u6c42\u3081\u308b\uff09<\/span>\r\n        <span class=\"n\">k_curr<\/span> <span class=\"o\">=<\/span> <span class=\"n\">W<\/span><span class=\"o\">.<\/span><span class=\"n\">shape<\/span><span class=\"p\">[<\/span><span class=\"mi\">1<\/span><span class=\"p\">]<\/span>\r\n        <span class=\"n\">H_test<\/span> <span class=\"o\">=<\/span> <span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">zeros<\/span><span class=\"p\">((<\/span><span class=\"n\">k_curr<\/span><span class=\"p\">,<\/span> <span class=\"n\">m_test<\/span><span class=\"p\">),<\/span> <span class=\"n\">dtype<\/span><span class=\"o\">=<\/span><span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">uint8<\/span><span class=\"p\">)<\/span>\r\n        <span class=\"n\">rng_test<\/span> <span class=\"o\">=<\/span> <span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">random<\/span><span class=\"o\">.<\/span><span class=\"n\">default_rng<\/span><span class=\"p\">(<\/span><span class=\"n\">seed<\/span> <span class=\"o\">+<\/span> <span class=\"n\">ep<\/span> <span class=\"o\">+<\/span> <span class=\"mi\">1000<\/span><span class=\"p\">)<\/span>\r\n        <span class=\"n\">total_test_elapsed_time<\/span> <span class=\"o\">=<\/span> <span class=\"mi\">0<\/span>\r\n        <span class=\"k\">for<\/span> <span class=\"n\">j<\/span> <span class=\"ow\">in<\/span> <span class=\"nb\">range<\/span><span class=\"p\">(<\/span><span class=\"n\">m_test<\/span><span class=\"p\">):<\/span>\r\n            <span class=\"n\">v<\/span> <span class=\"o\">=<\/span> <span class=\"n\">V_test_img<\/span><span class=\"p\">[:,<\/span> <span class=\"n\">j<\/span><span class=\"p\">]<\/span>\r\n            <span class=\"n\">b<\/span><span class=\"p\">,<\/span> <span class=\"n\">Q<\/span> <span class=\"o\">=<\/span> <span class=\"n\">precompute_qubo_terms<\/span><span class=\"p\">(<\/span><span class=\"n\">W<\/span><span class=\"p\">[:<\/span><span class=\"n\">n_img<\/span><span class=\"p\">,<\/span> <span class=\"p\">:],<\/span> <span class=\"n\">v<\/span><span class=\"o\">.<\/span><span class=\"n\">astype<\/span><span class=\"p\">(<\/span><span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">float32<\/span><span class=\"p\">),<\/span> <span class=\"n\">l1<\/span><span class=\"o\">=<\/span><span class=\"n\">l1<\/span><span class=\"p\">)<\/span>\r\n            <span class=\"n\">init_q<\/span> <span class=\"o\">=<\/span> <span class=\"n\">rng_test<\/span><span class=\"o\">.<\/span><span class=\"n\">integers<\/span><span class=\"p\">(<\/span><span class=\"mi\">0<\/span><span class=\"p\">,<\/span> <span class=\"mi\">2<\/span><span class=\"p\">,<\/span> <span class=\"n\">size<\/span><span class=\"o\">=<\/span><span class=\"n\">k_curr<\/span><span class=\"p\">,<\/span> <span class=\"n\">dtype<\/span><span class=\"o\">=<\/span><span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">uint8<\/span><span class=\"p\">)<\/span>\r\n            <span class=\"n\">q_opt<\/span><span class=\"p\">,<\/span> <span class=\"n\">_<\/span><span class=\"p\">,<\/span> <span class=\"n\">test_elapsed_time<\/span> <span class=\"o\">=<\/span> <span class=\"n\">simulated_annealing_qubo<\/span><span class=\"p\">(<\/span>\r\n                <span class=\"n\">b<\/span><span class=\"p\">,<\/span> <span class=\"n\">Q<\/span><span class=\"p\">,<\/span>\r\n                <span class=\"n\">num_reads<\/span><span class=\"o\">=<\/span><span class=\"n\">num_reads<\/span><span class=\"p\">,<\/span>\r\n                <span class=\"n\">seed<\/span><span class=\"o\">=<\/span><span class=\"nb\">int<\/span><span class=\"p\">(<\/span><span class=\"n\">rng_test<\/span><span class=\"o\">.<\/span><span class=\"n\">integers<\/span><span class=\"p\">(<\/span><span class=\"mi\">0<\/span><span class=\"p\">,<\/span> <span class=\"mi\">2<\/span><span class=\"o\">**<\/span><span class=\"mi\">31<\/span><span class=\"o\">-<\/span><span class=\"mi\">1<\/span><span class=\"p\">)),<\/span>\r\n                <span class=\"n\">sampler<\/span><span class=\"o\">=<\/span><span class=\"n\">sampler<\/span>\r\n            <span class=\"p\">)<\/span>\r\n            <span class=\"n\">H_test<\/span><span class=\"p\">[:,<\/span> <span class=\"n\">j<\/span><span class=\"p\">]<\/span> <span class=\"o\">=<\/span> <span class=\"n\">q_opt<\/span>\r\n            <span class=\"n\">total_test_elapsed_time<\/span> <span class=\"o\">+=<\/span> <span class=\"n\">test_elapsed_time<\/span>\r\n\r\n        <span class=\"c1\"># \u51fa\u529b\uff08\u5206\u985e\u5668\u90e8\uff09<\/span>\r\n        <span class=\"n\">W2<\/span> <span class=\"o\">=<\/span> <span class=\"n\">W<\/span><span class=\"p\">[<\/span><span class=\"n\">n_img<\/span><span class=\"p\">:,<\/span> <span class=\"p\">:]<\/span>\r\n        <span class=\"n\">U_test<\/span> <span class=\"o\">=<\/span> <span class=\"p\">(<\/span><span class=\"n\">W2<\/span> <span class=\"o\">@<\/span> <span class=\"n\">H_test<\/span><span class=\"o\">.<\/span><span class=\"n\">astype<\/span><span class=\"p\">(<\/span><span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">float32<\/span><span class=\"p\">))<\/span><span class=\"o\">.<\/span><span class=\"n\">astype<\/span><span class=\"p\">(<\/span><span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">float32<\/span><span class=\"p\">)<\/span>\r\n        <span class=\"n\">P<\/span> <span class=\"o\">=<\/span> <span class=\"n\">softmax<\/span><span class=\"p\">(<\/span><span class=\"n\">U_test<\/span><span class=\"p\">,<\/span> <span class=\"n\">axis<\/span><span class=\"o\">=<\/span><span class=\"mi\">0<\/span><span class=\"p\">)<\/span>\r\n        <span class=\"n\">y_pred<\/span> <span class=\"o\">=<\/span> <span class=\"n\">P<\/span><span class=\"o\">.<\/span><span class=\"n\">argmax<\/span><span class=\"p\">(<\/span><span class=\"n\">axis<\/span><span class=\"o\">=<\/span><span class=\"mi\">0<\/span><span class=\"p\">)<\/span>\r\n        <span class=\"n\">acc<\/span> <span class=\"o\">=<\/span> <span class=\"n\">accuracy_score<\/span><span class=\"p\">(<\/span><span class=\"n\">y_test<\/span><span class=\"p\">,<\/span> <span class=\"n\">y_pred<\/span><span class=\"p\">)<\/span>\r\n        <span class=\"n\">ce<\/span> <span class=\"o\">=<\/span> <span class=\"n\">log_loss<\/span><span class=\"p\">(<\/span><span class=\"n\">y_test<\/span><span class=\"p\">,<\/span> <span class=\"n\">P<\/span><span class=\"o\">.<\/span><span class=\"n\">T<\/span><span class=\"p\">,<\/span> <span class=\"n\">labels<\/span><span class=\"o\">=<\/span><span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">arange<\/span><span class=\"p\">(<\/span><span class=\"n\">n_classes<\/span><span class=\"p\">))<\/span>\r\n\r\n        <span class=\"n\">test_acc_list<\/span><span class=\"o\">.<\/span><span class=\"n\">append<\/span><span class=\"p\">(<\/span><span class=\"n\">acc<\/span><span class=\"p\">)<\/span>\r\n        <span class=\"n\">test_ce_list<\/span><span class=\"o\">.<\/span><span class=\"n\">append<\/span><span class=\"p\">(<\/span><span class=\"n\">ce<\/span><span class=\"p\">)<\/span>\r\n\r\n        <span class=\"n\">frob<\/span> <span class=\"o\">=<\/span> <span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">linalg<\/span><span class=\"o\">.<\/span><span class=\"n\">norm<\/span><span class=\"p\">((<\/span><span class=\"n\">V_train<\/span> <span class=\"o\">-<\/span> <span class=\"n\">W<\/span> <span class=\"o\">@<\/span> <span class=\"n\">H<\/span><span class=\"o\">.<\/span><span class=\"n\">astype<\/span><span class=\"p\">(<\/span><span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">float32<\/span><span class=\"p\">)),<\/span> <span class=\"s1\">'fro'<\/span><span class=\"p\">)<\/span>\r\n\r\n        <span class=\"nb\">print<\/span><span class=\"p\">(<\/span><span class=\"sa\">f<\/span><span class=\"s2\">\"[Epoch <\/span><span class=\"si\">{<\/span><span class=\"n\">ep<\/span><span class=\"si\">:<\/span><span class=\"s2\">02d<\/span><span class=\"si\">}<\/span><span class=\"s2\">] TestAcc=<\/span><span class=\"si\">{<\/span><span class=\"n\">acc<\/span><span class=\"o\">*<\/span><span class=\"mi\">100<\/span><span class=\"si\">:<\/span><span class=\"s2\">.2f<\/span><span class=\"si\">}<\/span><span class=\"s2\">%  CE=<\/span><span class=\"si\">{<\/span><span class=\"n\">ce<\/span><span class=\"si\">:<\/span><span class=\"s2\">.4f<\/span><span class=\"si\">}<\/span><span class=\"s2\">\"<\/span><span class=\"p\">)<\/span> <span class=\"c1\">#W_num={n_iter} W-elapsed={elapsed1:.1f}s H-elapsed={H_elapsed_time:.1f}s test-elapsed={total_test_elapsed_time:.1f}s\")<\/span>\r\n\r\n    <span class=\"k\">return<\/span> <span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">array<\/span><span class=\"p\">(<\/span><span class=\"n\">test_acc_list<\/span><span class=\"p\">),<\/span> <span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">array<\/span><span class=\"p\">(<\/span><span class=\"n\">test_ce_list<\/span><span class=\"p\">),<\/span> <span class=\"n\">W<\/span><span class=\"p\">,<\/span> <span class=\"n\">H<\/span><span class=\"p\">,<\/span> <span class=\"n\">y_pred<\/span><span class=\"p\">,<\/span> <span class=\"n\">H_test<\/span>\r\n<\/pre>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing text_cell rendered\">\n<div class=\"inner_cell\">\n<div class=\"text_cell_render border-box-sizing rendered_html\">\n<h3><span class=\"ez-toc-section\" id=\"FCNN%E3%81%AE%E5%AE%9F%E8%A3%85\"><\/span>FCNN\u306e\u5b9f\u88c5<span class=\"ez-toc-section-end\"><\/span><\/h3>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing text_cell rendered\">\n<div class=\"inner_cell\">\n<div class=\"text_cell_render border-box-sizing rendered_html\">\n<p>\u5b9f\u9a13\u3067NBMF\u3068\u6bd4\u8f03\u3059\u308b\u305f\u3081FCNN\u306b\u3088\u308b\u5b9f\u88c5\u3092\u884c\u3044\u307e\u3059\u3002<\/p>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing code_cell rendered\">\n<div class=\"input\">\n<div class=\"inner_cell\">\n<div class=\"input_area\">\n<div class=\"highlight hl-python\">\n<pre><span><\/span><span class=\"k\">class<\/span><span class=\"w\"> <\/span><span class=\"nc\">FCNN<\/span><span class=\"p\">(<\/span><span class=\"n\">nn<\/span><span class=\"o\">.<\/span><span class=\"n\">Module<\/span><span class=\"p\">):<\/span>\r\n    <span class=\"k\">def<\/span><span class=\"w\"> <\/span><span class=\"fm\">__init__<\/span><span class=\"p\">(<\/span><span class=\"bp\">self<\/span><span class=\"p\">,<\/span> <span class=\"n\">input_dim<\/span><span class=\"o\">=<\/span><span class=\"mi\">784<\/span><span class=\"p\">,<\/span> <span class=\"n\">hidden_dim<\/span><span class=\"o\">=<\/span><span class=\"mi\">40<\/span><span class=\"p\">,<\/span> <span class=\"n\">num_classes<\/span><span class=\"o\">=<\/span><span class=\"mi\">10<\/span><span class=\"p\">,<\/span> <span class=\"n\">dropout_rate<\/span><span class=\"o\">=<\/span><span class=\"mf\">0.2<\/span><span class=\"p\">):<\/span>\r\n        <span class=\"nb\">super<\/span><span class=\"p\">()<\/span><span class=\"o\">.<\/span><span class=\"fm\">__init__<\/span><span class=\"p\">()<\/span>\r\n        <span class=\"bp\">self<\/span><span class=\"o\">.<\/span><span class=\"n\">fc1<\/span> <span class=\"o\">=<\/span> <span class=\"n\">nn<\/span><span class=\"o\">.<\/span><span class=\"n\">Linear<\/span><span class=\"p\">(<\/span><span class=\"n\">input_dim<\/span><span class=\"p\">,<\/span> <span class=\"n\">hidden_dim<\/span><span class=\"p\">)<\/span>\r\n        <span class=\"bp\">self<\/span><span class=\"o\">.<\/span><span class=\"n\">relu<\/span> <span class=\"o\">=<\/span> <span class=\"n\">nn<\/span><span class=\"o\">.<\/span><span class=\"n\">ReLU<\/span><span class=\"p\">()<\/span>\r\n        <span class=\"bp\">self<\/span><span class=\"o\">.<\/span><span class=\"n\">dropout<\/span> <span class=\"o\">=<\/span> <span class=\"n\">nn<\/span><span class=\"o\">.<\/span><span class=\"n\">Dropout<\/span><span class=\"p\">(<\/span><span class=\"n\">p<\/span><span class=\"o\">=<\/span><span class=\"n\">dropout_rate<\/span><span class=\"p\">)<\/span>\r\n        <span class=\"bp\">self<\/span><span class=\"o\">.<\/span><span class=\"n\">fc2<\/span> <span class=\"o\">=<\/span> <span class=\"n\">nn<\/span><span class=\"o\">.<\/span><span class=\"n\">Linear<\/span><span class=\"p\">(<\/span><span class=\"n\">hidden_dim<\/span><span class=\"p\">,<\/span> <span class=\"n\">num_classes<\/span><span class=\"p\">)<\/span>\r\n\r\n    <span class=\"k\">def<\/span><span class=\"w\"> <\/span><span class=\"nf\">forward<\/span><span class=\"p\">(<\/span><span class=\"bp\">self<\/span><span class=\"p\">,<\/span> <span class=\"n\">x<\/span><span class=\"p\">):<\/span>\r\n        <span class=\"n\">x<\/span> <span class=\"o\">=<\/span> <span class=\"bp\">self<\/span><span class=\"o\">.<\/span><span class=\"n\">fc1<\/span><span class=\"p\">(<\/span><span class=\"n\">x<\/span><span class=\"p\">)<\/span>\r\n        <span class=\"n\">x<\/span> <span class=\"o\">=<\/span> <span class=\"bp\">self<\/span><span class=\"o\">.<\/span><span class=\"n\">relu<\/span><span class=\"p\">(<\/span><span class=\"n\">x<\/span><span class=\"p\">)<\/span>\r\n        <span class=\"n\">x<\/span> <span class=\"o\">=<\/span> <span class=\"bp\">self<\/span><span class=\"o\">.<\/span><span class=\"n\">dropout<\/span><span class=\"p\">(<\/span><span class=\"n\">x<\/span><span class=\"p\">)<\/span>\r\n        <span class=\"n\">x<\/span> <span class=\"o\">=<\/span> <span class=\"bp\">self<\/span><span class=\"o\">.<\/span><span class=\"n\">fc2<\/span><span class=\"p\">(<\/span><span class=\"n\">x<\/span><span class=\"p\">)<\/span>\r\n        <span class=\"k\">return<\/span> <span class=\"n\">x<\/span>\r\n<\/pre>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing code_cell rendered\">\n<div class=\"input\">\n<div class=\"inner_cell\">\n<div class=\"input_area\">\n<div class=\"highlight hl-python\">\n<pre><span><\/span><span class=\"k\">def<\/span><span class=\"w\"> <\/span><span class=\"nf\">train_fcnn_once<\/span><span class=\"p\">(<\/span><span class=\"n\">X_train<\/span><span class=\"p\">,<\/span> <span class=\"n\">y_train<\/span><span class=\"p\">,<\/span> <span class=\"n\">X_test<\/span><span class=\"p\">,<\/span> <span class=\"n\">y_test<\/span><span class=\"p\">,<\/span>\r\n                    <span class=\"n\">hidden_dim<\/span><span class=\"o\">=<\/span><span class=\"mi\">40<\/span><span class=\"p\">,<\/span> <span class=\"n\">lr<\/span><span class=\"o\">=<\/span><span class=\"mf\">0.001<\/span><span class=\"p\">,<\/span> <span class=\"n\">epochs<\/span><span class=\"o\">=<\/span><span class=\"mi\">10<\/span><span class=\"p\">,<\/span> <span class=\"n\">batch_size<\/span><span class=\"o\">=<\/span><span class=\"mi\">32<\/span><span class=\"p\">,<\/span> <span class=\"n\">seed<\/span><span class=\"o\">=<\/span><span class=\"mi\">0<\/span><span class=\"p\">):<\/span>\r\n<span class=\"w\">    <\/span><span class=\"sd\">\"\"\"<\/span>\r\n<span class=\"sd\">    FCNN \u306b\u3088\u308b\u5b66\u7fd2\u30921\u56de\u5206\u884c\u3046<\/span>\r\n<span class=\"sd\">    &lt;\u5165\u529b&gt;<\/span>\r\n<span class=\"sd\">    X_train\uff1a(m_train, n_features)<\/span>\r\n<span class=\"sd\">    y_train\uff1a(m_train,)<\/span>\r\n<span class=\"sd\">    X_test\uff1a(m_test, n_features)<\/span>\r\n<span class=\"sd\">    y_test\uff1a(m_test,)<\/span>\r\n<span class=\"sd\">    hidden_dim\uff1a\u96a0\u308c\u5c64\u306e\u6b21\u5143\u6570<\/span>\r\n<span class=\"sd\">    lr\uff1a\u5b66\u7fd2\u7387<\/span>\r\n<span class=\"sd\">    epochs\uff1a\u30a8\u30dd\u30c3\u30af\u6570<\/span>\r\n<span class=\"sd\">    batch_size\uff1a\u30d0\u30c3\u30c1\u30b5\u30a4\u30ba<\/span>\r\n<span class=\"sd\">    seed\uff1a\u4e71\u6570\u30b7\u30fc\u30c9<\/span>\r\n<span class=\"sd\">    &lt;\u51fa\u529b&gt;<\/span>\r\n<span class=\"sd\">    test_acc_list\uff1a(epochs,), \u30a8\u30dd\u30c3\u30af\u3054\u3068\u306e\u7cbe\u5ea6<\/span>\r\n<span class=\"sd\">    test_ce_list\uff1a(epochs,), \u30a8\u30dd\u30c3\u30af\u3054\u3068\u306e\u4ea4\u5dee\u30a8\u30f3\u30c8\u30ed\u30d4\u30fc\u8aa4\u5dee<\/span>\r\n<span class=\"sd\">    \"\"\"<\/span>\r\n    <span class=\"n\">torch<\/span><span class=\"o\">.<\/span><span class=\"n\">manual_seed<\/span><span class=\"p\">(<\/span><span class=\"n\">seed<\/span><span class=\"p\">)<\/span>\r\n    <span class=\"n\">device<\/span> <span class=\"o\">=<\/span> <span class=\"n\">torch<\/span><span class=\"o\">.<\/span><span class=\"n\">device<\/span><span class=\"p\">(<\/span><span class=\"s2\">\"cuda\"<\/span> <span class=\"k\">if<\/span> <span class=\"n\">torch<\/span><span class=\"o\">.<\/span><span class=\"n\">cuda<\/span><span class=\"o\">.<\/span><span class=\"n\">is_available<\/span><span class=\"p\">()<\/span> <span class=\"k\">else<\/span> <span class=\"s2\">\"cpu\"<\/span><span class=\"p\">)<\/span>\r\n\r\n    <span class=\"n\">X_train_t<\/span> <span class=\"o\">=<\/span> <span class=\"n\">torch<\/span><span class=\"o\">.<\/span><span class=\"n\">from_numpy<\/span><span class=\"p\">(<\/span><span class=\"n\">X_train<\/span><span class=\"p\">)<\/span><span class=\"o\">.<\/span><span class=\"n\">to<\/span><span class=\"p\">(<\/span><span class=\"n\">device<\/span><span class=\"p\">)<\/span>\r\n    <span class=\"n\">y_train_t<\/span> <span class=\"o\">=<\/span> <span class=\"n\">torch<\/span><span class=\"o\">.<\/span><span class=\"n\">from_numpy<\/span><span class=\"p\">(<\/span><span class=\"n\">y_train<\/span><span class=\"p\">)<\/span><span class=\"o\">.<\/span><span class=\"n\">to<\/span><span class=\"p\">(<\/span><span class=\"n\">device<\/span><span class=\"p\">)<\/span>\r\n    <span class=\"n\">X_test_t<\/span> <span class=\"o\">=<\/span> <span class=\"n\">torch<\/span><span class=\"o\">.<\/span><span class=\"n\">from_numpy<\/span><span class=\"p\">(<\/span><span class=\"n\">X_test<\/span><span class=\"p\">)<\/span><span class=\"o\">.<\/span><span class=\"n\">to<\/span><span class=\"p\">(<\/span><span class=\"n\">device<\/span><span class=\"p\">)<\/span>\r\n\r\n    <span class=\"n\">model<\/span> <span class=\"o\">=<\/span> <span class=\"n\">FCNN<\/span><span class=\"p\">(<\/span><span class=\"n\">hidden_dim<\/span><span class=\"o\">=<\/span><span class=\"n\">hidden_dim<\/span><span class=\"p\">)<\/span><span class=\"o\">.<\/span><span class=\"n\">to<\/span><span class=\"p\">(<\/span><span class=\"n\">device<\/span><span class=\"p\">)<\/span>\r\n    <span class=\"n\">criterion<\/span> <span class=\"o\">=<\/span> <span class=\"n\">nn<\/span><span class=\"o\">.<\/span><span class=\"n\">CrossEntropyLoss<\/span><span class=\"p\">()<\/span>\r\n    <span class=\"n\">optimizer<\/span> <span class=\"o\">=<\/span> <span class=\"n\">optim<\/span><span class=\"o\">.<\/span><span class=\"n\">Adam<\/span><span class=\"p\">(<\/span><span class=\"n\">model<\/span><span class=\"o\">.<\/span><span class=\"n\">parameters<\/span><span class=\"p\">(),<\/span> <span class=\"n\">lr<\/span><span class=\"o\">=<\/span><span class=\"n\">lr<\/span><span class=\"p\">)<\/span>\r\n\r\n    <span class=\"n\">test_acc_list<\/span> <span class=\"o\">=<\/span> <span class=\"p\">[]<\/span>\r\n    <span class=\"n\">test_ce_list<\/span> <span class=\"o\">=<\/span> <span class=\"p\">[]<\/span>\r\n\r\n    <span class=\"k\">for<\/span> <span class=\"n\">ep<\/span> <span class=\"ow\">in<\/span> <span class=\"nb\">range<\/span><span class=\"p\">(<\/span><span class=\"n\">epochs<\/span><span class=\"p\">):<\/span>\r\n        <span class=\"n\">model<\/span><span class=\"o\">.<\/span><span class=\"n\">train<\/span><span class=\"p\">()<\/span>\r\n        <span class=\"n\">perm<\/span> <span class=\"o\">=<\/span> <span class=\"n\">torch<\/span><span class=\"o\">.<\/span><span class=\"n\">randperm<\/span><span class=\"p\">(<\/span><span class=\"nb\">len<\/span><span class=\"p\">(<\/span><span class=\"n\">X_train_t<\/span><span class=\"p\">))<\/span>\r\n        <span class=\"n\">total_loss<\/span> <span class=\"o\">=<\/span> <span class=\"mf\">0.0<\/span>\r\n\r\n        <span class=\"k\">for<\/span> <span class=\"n\">i<\/span> <span class=\"ow\">in<\/span> <span class=\"nb\">range<\/span><span class=\"p\">(<\/span><span class=\"mi\">0<\/span><span class=\"p\">,<\/span> <span class=\"nb\">len<\/span><span class=\"p\">(<\/span><span class=\"n\">X_train_t<\/span><span class=\"p\">),<\/span> <span class=\"n\">batch_size<\/span><span class=\"p\">):<\/span>\r\n            <span class=\"n\">idx<\/span> <span class=\"o\">=<\/span> <span class=\"n\">perm<\/span><span class=\"p\">[<\/span><span class=\"n\">i<\/span><span class=\"p\">:<\/span><span class=\"n\">i<\/span><span class=\"o\">+<\/span><span class=\"n\">batch_size<\/span><span class=\"p\">]<\/span>\r\n            <span class=\"n\">xb<\/span><span class=\"p\">,<\/span> <span class=\"n\">yb<\/span> <span class=\"o\">=<\/span> <span class=\"n\">X_train_t<\/span><span class=\"p\">[<\/span><span class=\"n\">idx<\/span><span class=\"p\">],<\/span> <span class=\"n\">y_train_t<\/span><span class=\"p\">[<\/span><span class=\"n\">idx<\/span><span class=\"p\">]<\/span>\r\n\r\n            <span class=\"n\">optimizer<\/span><span class=\"o\">.<\/span><span class=\"n\">zero_grad<\/span><span class=\"p\">()<\/span>\r\n            <span class=\"n\">logits<\/span> <span class=\"o\">=<\/span> <span class=\"n\">model<\/span><span class=\"p\">(<\/span><span class=\"n\">xb<\/span><span class=\"p\">)<\/span>\r\n            <span class=\"n\">loss<\/span> <span class=\"o\">=<\/span> <span class=\"n\">criterion<\/span><span class=\"p\">(<\/span><span class=\"n\">logits<\/span><span class=\"p\">,<\/span> <span class=\"n\">yb<\/span><span class=\"p\">)<\/span>\r\n            <span class=\"n\">loss<\/span><span class=\"o\">.<\/span><span class=\"n\">backward<\/span><span class=\"p\">()<\/span>\r\n            <span class=\"n\">optimizer<\/span><span class=\"o\">.<\/span><span class=\"n\">step<\/span><span class=\"p\">()<\/span>\r\n            <span class=\"n\">total_loss<\/span> <span class=\"o\">+=<\/span> <span class=\"n\">loss<\/span><span class=\"o\">.<\/span><span class=\"n\">item<\/span><span class=\"p\">()<\/span>\r\n\r\n        <span class=\"c1\"># \u30c6\u30b9\u30c8\u8a55\u4fa1<\/span>\r\n        <span class=\"n\">model<\/span><span class=\"o\">.<\/span><span class=\"n\">eval<\/span><span class=\"p\">()<\/span>\r\n        <span class=\"k\">with<\/span> <span class=\"n\">torch<\/span><span class=\"o\">.<\/span><span class=\"n\">no_grad<\/span><span class=\"p\">():<\/span>\r\n            <span class=\"n\">logits_test<\/span> <span class=\"o\">=<\/span> <span class=\"n\">model<\/span><span class=\"p\">(<\/span><span class=\"n\">X_test_t<\/span><span class=\"p\">)<\/span>\r\n            <span class=\"n\">y_pred<\/span> <span class=\"o\">=<\/span> <span class=\"n\">logits_test<\/span><span class=\"o\">.<\/span><span class=\"n\">argmax<\/span><span class=\"p\">(<\/span><span class=\"n\">dim<\/span><span class=\"o\">=<\/span><span class=\"mi\">1<\/span><span class=\"p\">)<\/span><span class=\"o\">.<\/span><span class=\"n\">cpu<\/span><span class=\"p\">()<\/span><span class=\"o\">.<\/span><span class=\"n\">numpy<\/span><span class=\"p\">()<\/span>\r\n            <span class=\"n\">probs<\/span> <span class=\"o\">=<\/span> <span class=\"n\">torch<\/span><span class=\"o\">.<\/span><span class=\"n\">softmax<\/span><span class=\"p\">(<\/span><span class=\"n\">logits_test<\/span><span class=\"p\">,<\/span> <span class=\"n\">dim<\/span><span class=\"o\">=<\/span><span class=\"mi\">1<\/span><span class=\"p\">)<\/span><span class=\"o\">.<\/span><span class=\"n\">cpu<\/span><span class=\"p\">()<\/span><span class=\"o\">.<\/span><span class=\"n\">numpy<\/span><span class=\"p\">()<\/span>\r\n            <span class=\"n\">acc<\/span> <span class=\"o\">=<\/span> <span class=\"n\">accuracy_score<\/span><span class=\"p\">(<\/span><span class=\"n\">y_test<\/span><span class=\"p\">,<\/span> <span class=\"n\">y_pred<\/span><span class=\"p\">)<\/span>\r\n            <span class=\"n\">ce<\/span> <span class=\"o\">=<\/span> <span class=\"n\">log_loss<\/span><span class=\"p\">(<\/span><span class=\"n\">y_test<\/span><span class=\"p\">,<\/span> <span class=\"n\">probs<\/span><span class=\"p\">,<\/span> <span class=\"n\">labels<\/span><span class=\"o\">=<\/span><span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">arange<\/span><span class=\"p\">(<\/span><span class=\"mi\">10<\/span><span class=\"p\">))<\/span>\r\n        <span class=\"n\">test_acc_list<\/span><span class=\"o\">.<\/span><span class=\"n\">append<\/span><span class=\"p\">(<\/span><span class=\"n\">acc<\/span><span class=\"p\">)<\/span>\r\n        <span class=\"n\">test_ce_list<\/span><span class=\"o\">.<\/span><span class=\"n\">append<\/span><span class=\"p\">(<\/span><span class=\"n\">ce<\/span><span class=\"p\">)<\/span>\r\n\r\n    <span class=\"k\">return<\/span> <span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">array<\/span><span class=\"p\">(<\/span><span class=\"n\">test_acc_list<\/span><span class=\"p\">),<\/span> <span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">array<\/span><span class=\"p\">(<\/span><span class=\"n\">test_ce_list<\/span><span class=\"p\">)<\/span>\r\n<\/pre>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing text_cell rendered\">\n<div class=\"inner_cell\">\n<div class=\"text_cell_render border-box-sizing rendered_html\">\n<h3><span class=\"ez-toc-section\" id=\"%E7%B5%90%E6%9E%9C%E3%82%92%E5%8F%AF%E8%A6%96%E5%8C%96%E3%81%99%E3%82%8B%E3%81%9F%E3%82%81%E3%81%AE%E9%96%A2%E6%95%B0\"><\/span>\u7d50\u679c\u3092\u53ef\u8996\u5316\u3059\u308b\u305f\u3081\u306e\u95a2\u6570<span class=\"ez-toc-section-end\"><\/span><\/h3>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing text_cell rendered\">\n<div class=\"inner_cell\">\n<div class=\"text_cell_render border-box-sizing rendered_html\">\n<p>\u4ee5\u4e0b\u3067\u306f\u5b66\u7fd2\u30c7\u30fc\u30bf\u6570\u3001\u7279\u5fb4\u6570\u3001\u30a8\u30dd\u30c3\u30af\u6570\u306e\u305d\u308c\u305e\u308c\u306b\u5bfe\u3059\u308b\u7cbe\u5ea6\u3068\u4ea4\u5dee\u30a8\u30f3\u30c8\u30ed\u30d4\u30fc\u8aa4\u5dee\u3092\u30b0\u30e9\u30d5\u3068\u3057\u3066\u63cf\u753b\u3059\u308b\u95a2\u6570\u3092\u4f5c\u6210\u3057\u307e\u3059\u3002<\/p>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing code_cell rendered\">\n<div class=\"input\">\n<div class=\"inner_cell\">\n<div class=\"input_area\">\n<div class=\"highlight hl-python\">\n<pre><span><\/span><span class=\"k\">def<\/span><span class=\"w\"> <\/span><span class=\"nf\">plot_results_vs_train_size<\/span><span class=\"p\">(<\/span><span class=\"n\">train_sizes<\/span><span class=\"p\">,<\/span>\r\n                               <span class=\"n\">acc_mean_fcnn<\/span><span class=\"p\">,<\/span> <span class=\"n\">acc_std_fcnn<\/span><span class=\"p\">,<\/span> <span class=\"n\">ce_mean_fcnn<\/span><span class=\"p\">,<\/span> <span class=\"n\">ce_std_fcnn<\/span><span class=\"p\">,<\/span>\r\n                               <span class=\"n\">acc_mean_nbmf<\/span><span class=\"p\">,<\/span> <span class=\"n\">acc_std_nbmf<\/span><span class=\"p\">,<\/span> <span class=\"n\">ce_mean_nbmf<\/span><span class=\"p\">,<\/span> <span class=\"n\">ce_std_nbmf<\/span><span class=\"p\">):<\/span>\r\n<span class=\"w\">    <\/span><span class=\"sd\">\"\"\"<\/span>\r\n<span class=\"sd\">    FCNN\/NBMF \u306e\u5b66\u7fd2\u30c7\u30fc\u30bf\u306b\u5bfe\u3059\u308b\u7cbe\u5ea6\u30fb\u4ea4\u5dee\u30a8\u30f3\u30c8\u30ed\u30d4\u30fc\u8aa4\u5dee\u3092\u63cf\u753b\u3059\u308b<\/span>\r\n<span class=\"sd\">    &lt;\u5165\u529b&gt;<\/span>\r\n<span class=\"sd\">    train_sizes\uff1a\u5b66\u7fd2\u30c7\u30fc\u30bf\u6570\u306e\u30ea\u30b9\u30c8<\/span>\r\n<span class=\"sd\">    acc_mean_fcnn, acc_std_fcnn, ce_mean_fcnn, ce_std_fcnn\uff1aFCNN\u306e\u7cbe\u5ea6\u30fb\u4ea4\u5dee\u30a8\u30f3\u30c8\u30ed\u30d4\u30fc\u8aa4\u5dee<\/span>\r\n<span class=\"sd\">    acc_mean_nbmf, acc_std_nbmf, ce_mean_nbmf, ce_std_nbmf\uff1aNBMF\u306e\u7cbe\u5ea6\u30fb\u4ea4\u5dee\u30a8\u30f3\u30c8\u30ed\u30d4\u30fc\u8aa4\u5dee<\/span>\r\n<span class=\"sd\">    \"\"\"<\/span>\r\n    <span class=\"n\">fig<\/span><span class=\"p\">,<\/span> <span class=\"p\">(<\/span><span class=\"n\">ax1<\/span><span class=\"p\">,<\/span> <span class=\"n\">ax2<\/span><span class=\"p\">)<\/span> <span class=\"o\">=<\/span> <span class=\"n\">plt<\/span><span class=\"o\">.<\/span><span class=\"n\">subplots<\/span><span class=\"p\">(<\/span><span class=\"mi\">1<\/span><span class=\"p\">,<\/span> <span class=\"mi\">2<\/span><span class=\"p\">,<\/span> <span class=\"n\">figsize<\/span><span class=\"o\">=<\/span><span class=\"p\">(<\/span><span class=\"mi\">9<\/span><span class=\"p\">,<\/span><span class=\"mi\">4<\/span><span class=\"p\">))<\/span>\r\n\r\n    <span class=\"c1\"># (a) \u7cbe\u5ea6<\/span>\r\n    <span class=\"n\">ax1<\/span><span class=\"o\">.<\/span><span class=\"n\">errorbar<\/span><span class=\"p\">(<\/span><span class=\"n\">train_sizes<\/span><span class=\"p\">,<\/span> <span class=\"n\">acc_mean_fcnn<\/span><span class=\"o\">*<\/span><span class=\"mi\">100<\/span><span class=\"p\">,<\/span> <span class=\"n\">yerr<\/span><span class=\"o\">=<\/span><span class=\"n\">acc_std_fcnn<\/span><span class=\"o\">*<\/span><span class=\"mi\">100<\/span><span class=\"p\">,<\/span>\r\n                 <span class=\"n\">fmt<\/span><span class=\"o\">=<\/span><span class=\"s1\">'o-'<\/span><span class=\"p\">,<\/span> <span class=\"n\">color<\/span><span class=\"o\">=<\/span><span class=\"s1\">'red'<\/span><span class=\"p\">,<\/span> <span class=\"n\">label<\/span><span class=\"o\">=<\/span><span class=\"s1\">'FCNN'<\/span><span class=\"p\">,<\/span> <span class=\"n\">capsize<\/span><span class=\"o\">=<\/span><span class=\"mi\">3<\/span><span class=\"p\">)<\/span>\r\n    <span class=\"n\">ax1<\/span><span class=\"o\">.<\/span><span class=\"n\">errorbar<\/span><span class=\"p\">(<\/span><span class=\"n\">train_sizes<\/span><span class=\"p\">,<\/span> <span class=\"n\">acc_mean_nbmf<\/span><span class=\"o\">*<\/span><span class=\"mi\">100<\/span><span class=\"p\">,<\/span> <span class=\"n\">yerr<\/span><span class=\"o\">=<\/span><span class=\"n\">acc_std_nbmf<\/span><span class=\"o\">*<\/span><span class=\"mi\">100<\/span><span class=\"p\">,<\/span>\r\n                 <span class=\"n\">fmt<\/span><span class=\"o\">=<\/span><span class=\"s1\">'o-'<\/span><span class=\"p\">,<\/span> <span class=\"n\">color<\/span><span class=\"o\">=<\/span><span class=\"s1\">'blue'<\/span><span class=\"p\">,<\/span> <span class=\"n\">label<\/span><span class=\"o\">=<\/span><span class=\"s1\">'NBMF'<\/span><span class=\"p\">,<\/span> <span class=\"n\">capsize<\/span><span class=\"o\">=<\/span><span class=\"mi\">3<\/span><span class=\"p\">)<\/span>\r\n    <span class=\"n\">ax1<\/span><span class=\"o\">.<\/span><span class=\"n\">set_xlabel<\/span><span class=\"p\">(<\/span><span class=\"s2\">\"Training data size (m_train)\"<\/span><span class=\"p\">)<\/span>\r\n    <span class=\"n\">ax1<\/span><span class=\"o\">.<\/span><span class=\"n\">set_ylabel<\/span><span class=\"p\">(<\/span><span class=\"s2\">\"Test Accuracy (%)\"<\/span><span class=\"p\">)<\/span>\r\n    <span class=\"n\">ax1<\/span><span class=\"o\">.<\/span><span class=\"n\">set_title<\/span><span class=\"p\">(<\/span><span class=\"s2\">\"(a) Accuracy vs. training data size\"<\/span><span class=\"p\">)<\/span>\r\n    <span class=\"n\">ax1<\/span><span class=\"o\">.<\/span><span class=\"n\">grid<\/span><span class=\"p\">(<\/span><span class=\"kc\">True<\/span><span class=\"p\">,<\/span> <span class=\"n\">linestyle<\/span><span class=\"o\">=<\/span><span class=\"s2\">\"--\"<\/span><span class=\"p\">,<\/span> <span class=\"n\">alpha<\/span><span class=\"o\">=<\/span><span class=\"mf\">0.5<\/span><span class=\"p\">)<\/span>\r\n    <span class=\"n\">ax1<\/span><span class=\"o\">.<\/span><span class=\"n\">legend<\/span><span class=\"p\">()<\/span>\r\n\r\n    <span class=\"c1\"># (b) \u4ea4\u5dee\u30a8\u30f3\u30c8\u30ed\u30d4\u30fc\u8aa4\u5dee<\/span>\r\n    <span class=\"n\">ax2<\/span><span class=\"o\">.<\/span><span class=\"n\">errorbar<\/span><span class=\"p\">(<\/span><span class=\"n\">train_sizes<\/span><span class=\"p\">,<\/span> <span class=\"n\">ce_mean_fcnn<\/span><span class=\"p\">,<\/span> <span class=\"n\">yerr<\/span><span class=\"o\">=<\/span><span class=\"n\">ce_std_fcnn<\/span><span class=\"p\">,<\/span>\r\n                 <span class=\"n\">fmt<\/span><span class=\"o\">=<\/span><span class=\"s1\">'o-'<\/span><span class=\"p\">,<\/span> <span class=\"n\">color<\/span><span class=\"o\">=<\/span><span class=\"s1\">'red'<\/span><span class=\"p\">,<\/span> <span class=\"n\">label<\/span><span class=\"o\">=<\/span><span class=\"s1\">'FCNN'<\/span><span class=\"p\">,<\/span> <span class=\"n\">capsize<\/span><span class=\"o\">=<\/span><span class=\"mi\">3<\/span><span class=\"p\">)<\/span>\r\n    <span class=\"n\">ax2<\/span><span class=\"o\">.<\/span><span class=\"n\">errorbar<\/span><span class=\"p\">(<\/span><span class=\"n\">train_sizes<\/span><span class=\"p\">,<\/span> <span class=\"n\">ce_mean_nbmf<\/span><span class=\"p\">,<\/span> <span class=\"n\">yerr<\/span><span class=\"o\">=<\/span><span class=\"n\">ce_std_nbmf<\/span><span class=\"p\">,<\/span>\r\n                 <span class=\"n\">fmt<\/span><span class=\"o\">=<\/span><span class=\"s1\">'o-'<\/span><span class=\"p\">,<\/span> <span class=\"n\">color<\/span><span class=\"o\">=<\/span><span class=\"s1\">'blue'<\/span><span class=\"p\">,<\/span> <span class=\"n\">label<\/span><span class=\"o\">=<\/span><span class=\"s1\">'NBMF'<\/span><span class=\"p\">,<\/span> <span class=\"n\">capsize<\/span><span class=\"o\">=<\/span><span class=\"mi\">3<\/span><span class=\"p\">)<\/span>\r\n    <span class=\"n\">ax2<\/span><span class=\"o\">.<\/span><span class=\"n\">set_xlabel<\/span><span class=\"p\">(<\/span><span class=\"s2\">\"Training data size (m_train)\"<\/span><span class=\"p\">)<\/span>\r\n    <span class=\"n\">ax2<\/span><span class=\"o\">.<\/span><span class=\"n\">set_ylabel<\/span><span class=\"p\">(<\/span><span class=\"s2\">\"Cross-Entropy Loss\"<\/span><span class=\"p\">)<\/span>\r\n    <span class=\"n\">ax2<\/span><span class=\"o\">.<\/span><span class=\"n\">set_title<\/span><span class=\"p\">(<\/span><span class=\"s2\">\"(b) Cross-entropy vs. training data size\"<\/span><span class=\"p\">)<\/span>\r\n    <span class=\"n\">ax2<\/span><span class=\"o\">.<\/span><span class=\"n\">grid<\/span><span class=\"p\">(<\/span><span class=\"kc\">True<\/span><span class=\"p\">,<\/span> <span class=\"n\">linestyle<\/span><span class=\"o\">=<\/span><span class=\"s2\">\"--\"<\/span><span class=\"p\">,<\/span> <span class=\"n\">alpha<\/span><span class=\"o\">=<\/span><span class=\"mf\">0.5<\/span><span class=\"p\">)<\/span>\r\n    <span class=\"n\">ax2<\/span><span class=\"o\">.<\/span><span class=\"n\">legend<\/span><span class=\"p\">()<\/span>\r\n\r\n    <span class=\"n\">plt<\/span><span class=\"o\">.<\/span><span class=\"n\">tight_layout<\/span><span class=\"p\">()<\/span>\r\n    <span class=\"n\">plt<\/span><span class=\"o\">.<\/span><span class=\"n\">show<\/span><span class=\"p\">()<\/span>\r\n\r\n<span class=\"k\">def<\/span><span class=\"w\"> <\/span><span class=\"nf\">plot_results_vs_k<\/span><span class=\"p\">(<\/span><span class=\"n\">k_list<\/span><span class=\"p\">,<\/span>\r\n                      <span class=\"n\">acc_mean_fcnn<\/span><span class=\"p\">,<\/span> <span class=\"n\">acc_std_fcnn<\/span><span class=\"p\">,<\/span> <span class=\"n\">ce_mean_fcnn<\/span><span class=\"p\">,<\/span> <span class=\"n\">ce_std_fcnn<\/span><span class=\"p\">,<\/span>\r\n                      <span class=\"n\">acc_mean_nbmf<\/span><span class=\"p\">,<\/span> <span class=\"n\">acc_std_nbmf<\/span><span class=\"p\">,<\/span> <span class=\"n\">ce_mean_nbmf<\/span><span class=\"p\">,<\/span> <span class=\"n\">ce_std_nbmf<\/span><span class=\"p\">):<\/span>\r\n<span class=\"w\">    <\/span><span class=\"sd\">\"\"\"<\/span>\r\n<span class=\"sd\">    FCNN\/NBMF \u306e\u7279\u5fb4\u6570\u306b\u5bfe\u3059\u308b\u7cbe\u5ea6\u30fb\u4ea4\u5dee\u30a8\u30f3\u30c8\u30ed\u30d4\u30fc\u8aa4\u5dee\u3092\u63cf\u753b\u3059\u308b<\/span>\r\n<span class=\"sd\">    &lt;\u5165\u529b&gt;<\/span>\r\n<span class=\"sd\">    k_list\uff1a\u7279\u5fb4\u6570\u306e\u30ea\u30b9\u30c8<\/span>\r\n<span class=\"sd\">    acc_mean_fcnn, acc_std_fcnn, ce_mean_fcnn, ce_std_fcnn\uff1aFCNN\u306e\u7cbe\u5ea6\u30fb\u4ea4\u5dee\u30a8\u30f3\u30c8\u30ed\u30d4\u30fc\u8aa4\u5dee<\/span>\r\n<span class=\"sd\">    acc_mean_nbmf, acc_std_nbmf, ce_mean_nbmf, ce_std_nbmf\uff1aNBMF\u306e\u7cbe\u5ea6\u30fb\u4ea4\u5dee\u30a8\u30f3\u30c8\u30ed\u30d4\u30fc\u8aa4\u5dee<\/span>\r\n<span class=\"sd\">    \"\"\"<\/span>\r\n    <span class=\"n\">fig<\/span><span class=\"p\">,<\/span> <span class=\"p\">(<\/span><span class=\"n\">ax1<\/span><span class=\"p\">,<\/span> <span class=\"n\">ax2<\/span><span class=\"p\">)<\/span> <span class=\"o\">=<\/span> <span class=\"n\">plt<\/span><span class=\"o\">.<\/span><span class=\"n\">subplots<\/span><span class=\"p\">(<\/span><span class=\"mi\">1<\/span><span class=\"p\">,<\/span> <span class=\"mi\">2<\/span><span class=\"p\">,<\/span> <span class=\"n\">figsize<\/span><span class=\"o\">=<\/span><span class=\"p\">(<\/span><span class=\"mi\">9<\/span><span class=\"p\">,<\/span><span class=\"mi\">4<\/span><span class=\"p\">))<\/span>\r\n\r\n    <span class=\"c1\"># (a) \u7cbe\u5ea6<\/span>\r\n    <span class=\"n\">ax1<\/span><span class=\"o\">.<\/span><span class=\"n\">errorbar<\/span><span class=\"p\">(<\/span><span class=\"n\">k_list<\/span><span class=\"p\">,<\/span> <span class=\"n\">acc_mean_fcnn<\/span><span class=\"o\">*<\/span><span class=\"mi\">100<\/span><span class=\"p\">,<\/span> <span class=\"n\">yerr<\/span><span class=\"o\">=<\/span><span class=\"n\">acc_std_fcnn<\/span><span class=\"o\">*<\/span><span class=\"mi\">100<\/span><span class=\"p\">,<\/span>\r\n                 <span class=\"n\">fmt<\/span><span class=\"o\">=<\/span><span class=\"s1\">'o-'<\/span><span class=\"p\">,<\/span> <span class=\"n\">color<\/span><span class=\"o\">=<\/span><span class=\"s1\">'red'<\/span><span class=\"p\">,<\/span> <span class=\"n\">label<\/span><span class=\"o\">=<\/span><span class=\"s1\">'FCNN'<\/span><span class=\"p\">,<\/span> <span class=\"n\">capsize<\/span><span class=\"o\">=<\/span><span class=\"mi\">3<\/span><span class=\"p\">)<\/span>\r\n    <span class=\"n\">ax1<\/span><span class=\"o\">.<\/span><span class=\"n\">errorbar<\/span><span class=\"p\">(<\/span><span class=\"n\">k_list<\/span><span class=\"p\">,<\/span> <span class=\"n\">acc_mean_nbmf<\/span><span class=\"o\">*<\/span><span class=\"mi\">100<\/span><span class=\"p\">,<\/span> <span class=\"n\">yerr<\/span><span class=\"o\">=<\/span><span class=\"n\">acc_std_nbmf<\/span><span class=\"o\">*<\/span><span class=\"mi\">100<\/span><span class=\"p\">,<\/span>\r\n                 <span class=\"n\">fmt<\/span><span class=\"o\">=<\/span><span class=\"s1\">'o-'<\/span><span class=\"p\">,<\/span> <span class=\"n\">color<\/span><span class=\"o\">=<\/span><span class=\"s1\">'blue'<\/span><span class=\"p\">,<\/span> <span class=\"n\">label<\/span><span class=\"o\">=<\/span><span class=\"s1\">'NBMF'<\/span><span class=\"p\">,<\/span> <span class=\"n\">capsize<\/span><span class=\"o\">=<\/span><span class=\"mi\">3<\/span><span class=\"p\">)<\/span>\r\n    <span class=\"n\">ax1<\/span><span class=\"o\">.<\/span><span class=\"n\">set_xlabel<\/span><span class=\"p\">(<\/span><span class=\"s2\">\"k (hidden_dim \/ basis size)\"<\/span><span class=\"p\">)<\/span>\r\n    <span class=\"n\">ax1<\/span><span class=\"o\">.<\/span><span class=\"n\">set_ylabel<\/span><span class=\"p\">(<\/span><span class=\"s2\">\"Test Accuracy (%)\"<\/span><span class=\"p\">)<\/span>\r\n    <span class=\"n\">ax1<\/span><span class=\"o\">.<\/span><span class=\"n\">set_title<\/span><span class=\"p\">(<\/span><span class=\"s2\">\"(a) Accuracy vs. k\"<\/span><span class=\"p\">)<\/span>\r\n    <span class=\"n\">ax1<\/span><span class=\"o\">.<\/span><span class=\"n\">grid<\/span><span class=\"p\">(<\/span><span class=\"kc\">True<\/span><span class=\"p\">,<\/span> <span class=\"n\">linestyle<\/span><span class=\"o\">=<\/span><span class=\"s2\">\"--\"<\/span><span class=\"p\">,<\/span> <span class=\"n\">alpha<\/span><span class=\"o\">=<\/span><span class=\"mf\">0.5<\/span><span class=\"p\">)<\/span>\r\n    <span class=\"n\">ax1<\/span><span class=\"o\">.<\/span><span class=\"n\">legend<\/span><span class=\"p\">()<\/span>\r\n\r\n    <span class=\"c1\"># (b) \u4ea4\u5dee\u30a8\u30f3\u30c8\u30ed\u30d4\u30fc\u8aa4\u5dee<\/span>\r\n    <span class=\"n\">ax2<\/span><span class=\"o\">.<\/span><span class=\"n\">errorbar<\/span><span class=\"p\">(<\/span><span class=\"n\">k_list<\/span><span class=\"p\">,<\/span> <span class=\"n\">ce_mean_fcnn<\/span><span class=\"p\">,<\/span> <span class=\"n\">yerr<\/span><span class=\"o\">=<\/span><span class=\"n\">ce_std_fcnn<\/span><span class=\"p\">,<\/span>\r\n                 <span class=\"n\">fmt<\/span><span class=\"o\">=<\/span><span class=\"s1\">'o-'<\/span><span class=\"p\">,<\/span> <span class=\"n\">color<\/span><span class=\"o\">=<\/span><span class=\"s1\">'red'<\/span><span class=\"p\">,<\/span> <span class=\"n\">label<\/span><span class=\"o\">=<\/span><span class=\"s1\">'FCNN'<\/span><span class=\"p\">,<\/span> <span class=\"n\">capsize<\/span><span class=\"o\">=<\/span><span class=\"mi\">3<\/span><span class=\"p\">)<\/span>\r\n    <span class=\"n\">ax2<\/span><span class=\"o\">.<\/span><span class=\"n\">errorbar<\/span><span class=\"p\">(<\/span><span class=\"n\">k_list<\/span><span class=\"p\">,<\/span> <span class=\"n\">ce_mean_nbmf<\/span><span class=\"p\">,<\/span> <span class=\"n\">yerr<\/span><span class=\"o\">=<\/span><span class=\"n\">ce_std_nbmf<\/span><span class=\"p\">,<\/span>\r\n                 <span class=\"n\">fmt<\/span><span class=\"o\">=<\/span><span class=\"s1\">'o-'<\/span><span class=\"p\">,<\/span> <span class=\"n\">color<\/span><span class=\"o\">=<\/span><span class=\"s1\">'blue'<\/span><span class=\"p\">,<\/span> <span class=\"n\">label<\/span><span class=\"o\">=<\/span><span class=\"s1\">'NBMF'<\/span><span class=\"p\">,<\/span> <span class=\"n\">capsize<\/span><span class=\"o\">=<\/span><span class=\"mi\">3<\/span><span class=\"p\">)<\/span>\r\n    <span class=\"n\">ax2<\/span><span class=\"o\">.<\/span><span class=\"n\">set_xlabel<\/span><span class=\"p\">(<\/span><span class=\"s2\">\"k (hidden_dim \/ basis size)\"<\/span><span class=\"p\">)<\/span>\r\n    <span class=\"n\">ax2<\/span><span class=\"o\">.<\/span><span class=\"n\">set_ylabel<\/span><span class=\"p\">(<\/span><span class=\"s2\">\"Cross-Entropy Loss\"<\/span><span class=\"p\">)<\/span>\r\n    <span class=\"n\">ax2<\/span><span class=\"o\">.<\/span><span class=\"n\">set_title<\/span><span class=\"p\">(<\/span><span class=\"s2\">\"(b) Cross-entropy vs. k\"<\/span><span class=\"p\">)<\/span>\r\n    <span class=\"n\">ax2<\/span><span class=\"o\">.<\/span><span class=\"n\">grid<\/span><span class=\"p\">(<\/span><span class=\"kc\">True<\/span><span class=\"p\">,<\/span> <span class=\"n\">linestyle<\/span><span class=\"o\">=<\/span><span class=\"s2\">\"--\"<\/span><span class=\"p\">,<\/span> <span class=\"n\">alpha<\/span><span class=\"o\">=<\/span><span class=\"mf\">0.5<\/span><span class=\"p\">)<\/span>\r\n    <span class=\"n\">ax2<\/span><span class=\"o\">.<\/span><span class=\"n\">legend<\/span><span class=\"p\">()<\/span>\r\n\r\n    <span class=\"n\">plt<\/span><span class=\"o\">.<\/span><span class=\"n\">tight_layout<\/span><span class=\"p\">()<\/span>\r\n    <span class=\"n\">plt<\/span><span class=\"o\">.<\/span><span class=\"n\">show<\/span><span class=\"p\">()<\/span>\r\n\r\n<span class=\"k\">def<\/span><span class=\"w\"> <\/span><span class=\"nf\">plot_learning_curves<\/span><span class=\"p\">(<\/span><span class=\"n\">fcnn_epochs<\/span><span class=\"p\">,<\/span> <span class=\"n\">nbmf_epochs<\/span><span class=\"p\">,<\/span> <span class=\"n\">acc_mean_fcnn<\/span><span class=\"p\">,<\/span> <span class=\"n\">acc_std_fcnn<\/span><span class=\"p\">,<\/span> <span class=\"n\">ce_mean_fcnn<\/span><span class=\"p\">,<\/span> <span class=\"n\">ce_std_fcnn<\/span><span class=\"p\">,<\/span>\r\n                         <span class=\"n\">acc_mean_nbmf<\/span><span class=\"p\">,<\/span> <span class=\"n\">acc_std_nbmf<\/span><span class=\"p\">,<\/span> <span class=\"n\">ce_mean_nbmf<\/span><span class=\"p\">,<\/span> <span class=\"n\">ce_std_nbmf<\/span><span class=\"p\">):<\/span>\r\n<span class=\"w\">    <\/span><span class=\"sd\">\"\"\"<\/span>\r\n<span class=\"sd\">    FCNN\/NBMF \u306e\u30a8\u30dd\u30c3\u30af\u6570\u306b\u5bfe\u3059\u308b\u7cbe\u5ea6\u30fb\u4ea4\u5dee\u30a8\u30f3\u30c8\u30ed\u30d4\u30fc\u8aa4\u5dee\u3092\u63cf\u753b\u3059\u308b<\/span>\r\n<span class=\"sd\">    &lt;\u5165\u529b&gt;<\/span>\r\n<span class=\"sd\">    fcnn_epochs\uff1aFCNN\u306e\u30a8\u30dd\u30c3\u30af\u6570<\/span>\r\n<span class=\"sd\">    nbmf_epochs\uff1aNBMF\u306e\u30a8\u30dd\u30c3\u30af\u6570<\/span>\r\n<span class=\"sd\">    acc_mean_fcnn, acc_std_fcnn, ce_mean_fcnn, ce_std_fcnn\uff1aFCNN\u306e\u7cbe\u5ea6\u30fb\u4ea4\u5dee\u30a8\u30f3\u30c8\u30ed\u30d4\u30fc\u8aa4\u5dee<\/span>\r\n<span class=\"sd\">    acc_mean_nbmf, acc_std_nbmf, ce_mean_nbmf, ce_std_nbmf\uff1aNBMF\u306e\u7cbe\u5ea6\u30fb\u4ea4\u5dee\u30a8\u30f3\u30c8\u30ed\u30d4\u30fc\u8aa4\u5dee<\/span>\r\n<span class=\"sd\">    \"\"\"<\/span>\r\n    <span class=\"n\">fig<\/span><span class=\"p\">,<\/span> <span class=\"p\">(<\/span><span class=\"n\">ax1<\/span><span class=\"p\">,<\/span> <span class=\"n\">ax2<\/span><span class=\"p\">)<\/span> <span class=\"o\">=<\/span> <span class=\"n\">plt<\/span><span class=\"o\">.<\/span><span class=\"n\">subplots<\/span><span class=\"p\">(<\/span><span class=\"mi\">1<\/span><span class=\"p\">,<\/span> <span class=\"mi\">2<\/span><span class=\"p\">,<\/span> <span class=\"n\">figsize<\/span><span class=\"o\">=<\/span><span class=\"p\">(<\/span><span class=\"mi\">9<\/span><span class=\"p\">,<\/span><span class=\"mi\">4<\/span><span class=\"p\">))<\/span>\r\n    <span class=\"n\">x1<\/span> <span class=\"o\">=<\/span> <span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">arange<\/span><span class=\"p\">(<\/span><span class=\"mi\">1<\/span><span class=\"p\">,<\/span> <span class=\"n\">fcnn_epochs<\/span><span class=\"o\">+<\/span><span class=\"mi\">1<\/span><span class=\"p\">)<\/span>\r\n    <span class=\"n\">x2<\/span> <span class=\"o\">=<\/span> <span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">arange<\/span><span class=\"p\">(<\/span><span class=\"mi\">1<\/span><span class=\"p\">,<\/span> <span class=\"n\">nbmf_epochs<\/span><span class=\"o\">+<\/span><span class=\"mi\">1<\/span><span class=\"p\">)<\/span>\r\n\r\n    <span class=\"c1\"># (a) \u7cbe\u5ea6<\/span>\r\n    <span class=\"n\">ax1<\/span><span class=\"o\">.<\/span><span class=\"n\">plot<\/span><span class=\"p\">(<\/span><span class=\"n\">x1<\/span><span class=\"p\">,<\/span> <span class=\"n\">acc_mean_fcnn<\/span><span class=\"o\">*<\/span><span class=\"mi\">100<\/span><span class=\"p\">,<\/span> <span class=\"n\">label<\/span><span class=\"o\">=<\/span><span class=\"s1\">'FCNN'<\/span><span class=\"p\">,<\/span> <span class=\"n\">color<\/span><span class=\"o\">=<\/span><span class=\"s1\">'red'<\/span><span class=\"p\">)<\/span>\r\n\r\n    <span class=\"n\">ax1<\/span><span class=\"o\">.<\/span><span class=\"n\">plot<\/span><span class=\"p\">(<\/span><span class=\"n\">x2<\/span><span class=\"p\">,<\/span> <span class=\"n\">acc_mean_nbmf<\/span><span class=\"o\">*<\/span><span class=\"mi\">100<\/span><span class=\"p\">,<\/span> <span class=\"n\">label<\/span><span class=\"o\">=<\/span><span class=\"s1\">'NBMF'<\/span><span class=\"p\">,<\/span> <span class=\"n\">color<\/span><span class=\"o\">=<\/span><span class=\"s1\">'blue'<\/span><span class=\"p\">)<\/span>\r\n\r\n    <span class=\"n\">ax1<\/span><span class=\"o\">.<\/span><span class=\"n\">set_xscale<\/span><span class=\"p\">(<\/span><span class=\"s2\">\"log\"<\/span><span class=\"p\">)<\/span>\r\n    <span class=\"n\">ax1<\/span><span class=\"o\">.<\/span><span class=\"n\">set_xlabel<\/span><span class=\"p\">(<\/span><span class=\"s2\">\"Epoch (log scale)\"<\/span><span class=\"p\">)<\/span>\r\n    <span class=\"n\">ax1<\/span><span class=\"o\">.<\/span><span class=\"n\">set_ylabel<\/span><span class=\"p\">(<\/span><span class=\"s2\">\"Average Test Accuracy (%)\"<\/span><span class=\"p\">)<\/span>\r\n    <span class=\"n\">ax1<\/span><span class=\"o\">.<\/span><span class=\"n\">set_title<\/span><span class=\"p\">(<\/span><span class=\"s2\">\"(a) Average test accuracy vs. epochs\"<\/span><span class=\"p\">)<\/span>\r\n    <span class=\"n\">ax1<\/span><span class=\"o\">.<\/span><span class=\"n\">legend<\/span><span class=\"p\">()<\/span>\r\n    <span class=\"n\">ax1<\/span><span class=\"o\">.<\/span><span class=\"n\">grid<\/span><span class=\"p\">(<\/span><span class=\"kc\">True<\/span><span class=\"p\">,<\/span> <span class=\"n\">which<\/span><span class=\"o\">=<\/span><span class=\"s2\">\"both\"<\/span><span class=\"p\">,<\/span> <span class=\"n\">linestyle<\/span><span class=\"o\">=<\/span><span class=\"s2\">\"--\"<\/span><span class=\"p\">,<\/span> <span class=\"n\">alpha<\/span><span class=\"o\">=<\/span><span class=\"mf\">0.5<\/span><span class=\"p\">)<\/span>\r\n\r\n    <span class=\"c1\"># (b) \u4ea4\u5dee\u30a8\u30f3\u30c8\u30ed\u30d4\u30fc\u8aa4\u5dee<\/span>\r\n    <span class=\"n\">ax2<\/span><span class=\"o\">.<\/span><span class=\"n\">plot<\/span><span class=\"p\">(<\/span><span class=\"n\">x1<\/span><span class=\"p\">,<\/span> <span class=\"n\">ce_mean_fcnn<\/span><span class=\"p\">,<\/span> <span class=\"n\">label<\/span><span class=\"o\">=<\/span><span class=\"s1\">'FCNN'<\/span><span class=\"p\">,<\/span> <span class=\"n\">color<\/span><span class=\"o\">=<\/span><span class=\"s1\">'red'<\/span><span class=\"p\">)<\/span>\r\n\r\n    <span class=\"n\">ax2<\/span><span class=\"o\">.<\/span><span class=\"n\">plot<\/span><span class=\"p\">(<\/span><span class=\"n\">x2<\/span><span class=\"p\">,<\/span> <span class=\"n\">ce_mean_nbmf<\/span><span class=\"p\">,<\/span> <span class=\"n\">label<\/span><span class=\"o\">=<\/span><span class=\"s1\">'NBMF'<\/span><span class=\"p\">,<\/span> <span class=\"n\">color<\/span><span class=\"o\">=<\/span><span class=\"s1\">'blue'<\/span><span class=\"p\">)<\/span>\r\n\r\n    <span class=\"n\">ax2<\/span><span class=\"o\">.<\/span><span class=\"n\">set_xscale<\/span><span class=\"p\">(<\/span><span class=\"s2\">\"log\"<\/span><span class=\"p\">)<\/span>\r\n    <span class=\"n\">ax2<\/span><span class=\"o\">.<\/span><span class=\"n\">set_xlabel<\/span><span class=\"p\">(<\/span><span class=\"s2\">\"Epoch (log scale)\"<\/span><span class=\"p\">)<\/span>\r\n    <span class=\"n\">ax2<\/span><span class=\"o\">.<\/span><span class=\"n\">set_ylabel<\/span><span class=\"p\">(<\/span><span class=\"s2\">\"Cross-Entropy Loss\"<\/span><span class=\"p\">)<\/span>\r\n    <span class=\"n\">ax2<\/span><span class=\"o\">.<\/span><span class=\"n\">set_title<\/span><span class=\"p\">(<\/span><span class=\"s2\">\"(b) Cross-entropy loss vs. epochs\"<\/span><span class=\"p\">)<\/span>\r\n    <span class=\"n\">ax2<\/span><span class=\"o\">.<\/span><span class=\"n\">legend<\/span><span class=\"p\">()<\/span>\r\n    <span class=\"n\">ax2<\/span><span class=\"o\">.<\/span><span class=\"n\">grid<\/span><span class=\"p\">(<\/span><span class=\"kc\">True<\/span><span class=\"p\">,<\/span> <span class=\"n\">which<\/span><span class=\"o\">=<\/span><span class=\"s2\">\"both\"<\/span><span class=\"p\">,<\/span> <span class=\"n\">linestyle<\/span><span class=\"o\">=<\/span><span class=\"s2\">\"--\"<\/span><span class=\"p\">,<\/span> <span class=\"n\">alpha<\/span><span class=\"o\">=<\/span><span class=\"mf\">0.5<\/span><span class=\"p\">)<\/span>\r\n\r\n    <span class=\"n\">plt<\/span><span class=\"o\">.<\/span><span class=\"n\">tight_layout<\/span><span class=\"p\">()<\/span>\r\n    <span class=\"n\">plt<\/span><span class=\"o\">.<\/span><span class=\"n\">show<\/span><span class=\"p\">()<\/span>\r\n<\/pre>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing text_cell rendered\">\n<div class=\"inner_cell\">\n<div class=\"text_cell_render border-box-sizing rendered_html\">\n<h2><span class=\"ez-toc-section\" id=\"%E5%AE%9F%E9%A8%93\"><\/span>\u5b9f\u9a13<span class=\"ez-toc-section-end\"><\/span><\/h2>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing text_cell rendered\">\n<div class=\"inner_cell\">\n<div class=\"text_cell_render border-box-sizing rendered_html\">\n<p>\u5143\u8ad6\u6587\u3067\u306e\u5b9f\u9a13\u3068\u540c\u69d8\u306b\u3001\u672c\u8a18\u4e8b\u3067\u306f\u5b66\u7fd2\u30c7\u30fc\u30bf\u3001\u7279\u5fb4\u6570\u3001\u30a8\u30dd\u30c3\u30af\u6570\u3092\u305d\u308c\u305e\u308c\u5909\u5316\u3055\u305b\u305f\u5834\u5408\u306e\u7cbe\u5ea6\u3068\u4ea4\u5dee\u30a8\u30f3\u30c8\u30ed\u30d4\u30fc\u8aa4\u5dee\u3067NBMF\u3068FCNN\u306e\u6bd4\u8f03\u5b9f\u9a13\u3092\u884c\u3044\u307e\u3059\u3002<\/p>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing text_cell rendered\">\n<div class=\"inner_cell\">\n<div class=\"text_cell_render border-box-sizing rendered_html\">\n<h3><span class=\"ez-toc-section\" id=\"%E5%AD%A6%E7%BF%92%E3%83%87%E3%83%BC%E3%82%BF%E6%95%B0%E3%81%AB%E5%AF%BE%E3%81%99%E3%82%8B%E7%B2%BE%E5%BA%A6%E3%81%A8%E4%BA%A4%E5%B7%AE%E3%82%A8%E3%83%B3%E3%83%88%E3%83%AD%E3%83%94%E3%83%BC%E8%AA%A4%E5%B7%AE%E3%81%A7NBMF%E3%81%A8FCNN%E3%81%AE%E6%AF%94%E8%BC%83%E3%82%92%E8%A1%8C%E3%81%86%E5%AE%9F%E9%A8%93\"><\/span>\u5b66\u7fd2\u30c7\u30fc\u30bf\u6570\u306b\u5bfe\u3059\u308b\u7cbe\u5ea6\u3068\u4ea4\u5dee\u30a8\u30f3\u30c8\u30ed\u30d4\u30fc\u8aa4\u5dee\u3067NBMF\u3068FCNN\u306e\u6bd4\u8f03\u3092\u884c\u3046\u5b9f\u9a13<span class=\"ez-toc-section-end\"><\/span><\/h3>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing text_cell rendered\">\n<div class=\"inner_cell\">\n<div class=\"text_cell_render border-box-sizing rendered_html\">\n<p>\u307e\u305a\u3001\u5b66\u7fd2\u30c7\u30fc\u30bf\u6570\u309250\u304b\u3089350\u307e\u306725\u305a\u3064\u5909\u5316\u3055\u305b\u305f\u5834\u5408\u306e\u7cbe\u5ea6\u3068\u4ea4\u5dee\u30a8\u30f3\u30c8\u30ed\u30d4\u30fc\u8aa4\u5dee\u306b\u3064\u3044\u3066, NBMF\u3068FCNN\u306e\u6027\u80fd\u3092\u6bd4\u8f03\u3059\u308b\u5b9f\u9a13\u3092\u884c\u3044\u307e\u3059\u3002\u5404\u7a2e\u30d1\u30e9\u30e1\u30fc\u30bf\u306f\u8ad6\u6587\u306b\u57fa\u3065\u3044\u3066\u4ee5\u4e0b\u306e\u3088\u3046\u306b\u8a2d\u5b9a\u3057\u307e\u3059\u3002<\/p>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing code_cell rendered\">\n<div class=\"input\">\n<div class=\"inner_cell\">\n<div class=\"input_area\">\n<div class=\"highlight hl-python\">\n<pre><span><\/span><span class=\"c1\"># \u5171\u901a\u30d1\u30e9\u30e1\u30fc\u30bf<\/span>\r\n<span class=\"n\">epochs<\/span> <span class=\"o\">=<\/span> <span class=\"mi\">10<\/span>\r\n<span class=\"n\">num_repeats<\/span> <span class=\"o\">=<\/span> <span class=\"mi\">3<\/span>\r\n<span class=\"n\">m_test<\/span> <span class=\"o\">=<\/span> <span class=\"mi\">500<\/span>\r\n<span class=\"n\">seed<\/span> <span class=\"o\">=<\/span> <span class=\"mi\">0<\/span>\r\n<span class=\"n\">train_sizes<\/span> <span class=\"o\">=<\/span> <span class=\"nb\">list<\/span><span class=\"p\">(<\/span><span class=\"nb\">range<\/span><span class=\"p\">(<\/span><span class=\"mi\">50<\/span><span class=\"p\">,<\/span> <span class=\"mi\">351<\/span><span class=\"p\">,<\/span> <span class=\"mi\">25<\/span><span class=\"p\">))<\/span>\r\n\r\n<span class=\"c1\"># --- FCNN\u30d1\u30e9\u30e1\u30fc\u30bf ---<\/span>\r\n<span class=\"n\">hidden_dim<\/span> <span class=\"o\">=<\/span> <span class=\"mi\">40<\/span>\r\n<span class=\"n\">lr<\/span> <span class=\"o\">=<\/span> <span class=\"mf\">0.0002<\/span>\r\n<span class=\"n\">batch_size<\/span> <span class=\"o\">=<\/span> <span class=\"mi\">32<\/span>\r\n\r\n<span class=\"c1\"># --- NBMF\u30d1\u30e9\u30e1\u30fc\u30bf ---<\/span>\r\n<span class=\"n\">k<\/span> <span class=\"o\">=<\/span> <span class=\"mi\">40<\/span>\r\n<span class=\"n\">g<\/span> <span class=\"o\">=<\/span> <span class=\"mf\">9.0<\/span>\r\n<span class=\"n\">alpha<\/span> <span class=\"o\">=<\/span> <span class=\"mf\">1e-4<\/span>\r\n<span class=\"n\">lr_W<\/span> <span class=\"o\">=<\/span> <span class=\"mf\">0.01<\/span>\r\n<span class=\"n\">l1<\/span> <span class=\"o\">=<\/span> <span class=\"mi\">0<\/span>\r\n<span class=\"c1\"># sampler = oj.SASampler()<\/span>\r\n<span class=\"n\">sampler<\/span> <span class=\"o\">=<\/span> <span class=\"n\">neal<\/span><span class=\"o\">.<\/span><span class=\"n\">SimulatedAnnealingSampler<\/span><span class=\"p\">()<\/span>\r\n<\/pre>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing code_cell rendered\">\n<div class=\"input\">\n<div class=\"inner_cell\">\n<div class=\"input_area\">\n<div class=\"highlight hl-python\">\n<pre><span><\/span><span class=\"n\">acc_mean_fcnn_list<\/span><span class=\"p\">,<\/span> <span class=\"n\">acc_std_fcnn_list<\/span> <span class=\"o\">=<\/span> <span class=\"p\">[],<\/span> <span class=\"p\">[]<\/span>\r\n<span class=\"n\">ce_mean_fcnn_list<\/span><span class=\"p\">,<\/span> <span class=\"n\">ce_std_fcnn_list<\/span> <span class=\"o\">=<\/span> <span class=\"p\">[],<\/span> <span class=\"p\">[]<\/span>\r\n<span class=\"n\">acc_mean_nbmf_list<\/span><span class=\"p\">,<\/span> <span class=\"n\">acc_std_nbmf_list<\/span> <span class=\"o\">=<\/span> <span class=\"p\">[],<\/span> <span class=\"p\">[]<\/span>\r\n<span class=\"n\">ce_mean_nbmf_list<\/span><span class=\"p\">,<\/span> <span class=\"n\">ce_std_nbmf_list<\/span> <span class=\"o\">=<\/span> <span class=\"p\">[],<\/span> <span class=\"p\">[]<\/span>\r\n\r\n<span class=\"k\">for<\/span> <span class=\"n\">m_train<\/span> <span class=\"ow\">in<\/span> <span class=\"n\">train_sizes<\/span><span class=\"p\">:<\/span>\r\n    <span class=\"nb\">print<\/span><span class=\"p\">(<\/span><span class=\"sa\">f<\/span><span class=\"s2\">\"<\/span><span class=\"se\">\\n<\/span><span class=\"s2\">--- m_train = <\/span><span class=\"si\">{<\/span><span class=\"n\">m_train<\/span><span class=\"si\">}<\/span><span class=\"s2\"> ---\"<\/span><span class=\"p\">)<\/span>\r\n\r\n    <span class=\"c1\"># \u30c7\u30fc\u30bf\u8aad\u307f\u8fbc\u307f<\/span>\r\n    <span class=\"n\">X_train<\/span><span class=\"p\">,<\/span> <span class=\"n\">y_train<\/span><span class=\"p\">,<\/span> <span class=\"n\">X_test<\/span><span class=\"p\">,<\/span> <span class=\"n\">y_test<\/span> <span class=\"o\">=<\/span> <span class=\"n\">load_mnist_data<\/span><span class=\"p\">(<\/span><span class=\"n\">m_train<\/span><span class=\"p\">,<\/span> <span class=\"n\">m_test<\/span><span class=\"p\">,<\/span> <span class=\"n\">seed<\/span><span class=\"o\">=<\/span><span class=\"n\">seed<\/span><span class=\"p\">)<\/span>\r\n\r\n    <span class=\"c1\"># ---- FCNN ----<\/span>\r\n    <span class=\"n\">accs_fcnn<\/span><span class=\"p\">,<\/span> <span class=\"n\">ces_fcnn<\/span> <span class=\"o\">=<\/span> <span class=\"p\">[],<\/span> <span class=\"p\">[]<\/span>\r\n    <span class=\"k\">for<\/span> <span class=\"n\">s<\/span> <span class=\"ow\">in<\/span> <span class=\"nb\">range<\/span><span class=\"p\">(<\/span><span class=\"n\">num_repeats<\/span><span class=\"p\">):<\/span>\r\n        <span class=\"n\">acc_fcnn<\/span><span class=\"p\">,<\/span> <span class=\"n\">ce_fcnn<\/span> <span class=\"o\">=<\/span> <span class=\"n\">train_fcnn_once<\/span><span class=\"p\">(<\/span>\r\n            <span class=\"n\">X_train<\/span><span class=\"p\">,<\/span> <span class=\"n\">y_train<\/span><span class=\"p\">,<\/span> <span class=\"n\">X_test<\/span><span class=\"p\">,<\/span> <span class=\"n\">y_test<\/span><span class=\"p\">,<\/span>\r\n            <span class=\"n\">hidden_dim<\/span><span class=\"o\">=<\/span><span class=\"n\">hidden_dim<\/span><span class=\"p\">,<\/span> <span class=\"n\">lr<\/span><span class=\"o\">=<\/span><span class=\"n\">lr<\/span><span class=\"p\">,<\/span> <span class=\"n\">epochs<\/span><span class=\"o\">=<\/span><span class=\"n\">epochs<\/span><span class=\"p\">,<\/span>\r\n            <span class=\"n\">batch_size<\/span><span class=\"o\">=<\/span><span class=\"n\">batch_size<\/span><span class=\"p\">,<\/span> <span class=\"n\">seed<\/span><span class=\"o\">=<\/span><span class=\"n\">s<\/span>\r\n        <span class=\"p\">)<\/span>\r\n        <span class=\"n\">accs_fcnn<\/span><span class=\"o\">.<\/span><span class=\"n\">append<\/span><span class=\"p\">(<\/span><span class=\"n\">acc_fcnn<\/span><span class=\"p\">[<\/span><span class=\"o\">-<\/span><span class=\"mi\">1<\/span><span class=\"p\">])<\/span>\r\n        <span class=\"n\">ces_fcnn<\/span><span class=\"o\">.<\/span><span class=\"n\">append<\/span><span class=\"p\">(<\/span><span class=\"n\">ce_fcnn<\/span><span class=\"p\">[<\/span><span class=\"o\">-<\/span><span class=\"mi\">1<\/span><span class=\"p\">])<\/span>\r\n\r\n    <span class=\"n\">acc_mean_fcnn_list<\/span><span class=\"o\">.<\/span><span class=\"n\">append<\/span><span class=\"p\">(<\/span><span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">mean<\/span><span class=\"p\">(<\/span><span class=\"n\">accs_fcnn<\/span><span class=\"p\">))<\/span>\r\n    <span class=\"n\">acc_std_fcnn_list<\/span><span class=\"o\">.<\/span><span class=\"n\">append<\/span><span class=\"p\">(<\/span><span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">std<\/span><span class=\"p\">(<\/span><span class=\"n\">accs_fcnn<\/span><span class=\"p\">))<\/span>\r\n    <span class=\"n\">ce_mean_fcnn_list<\/span><span class=\"o\">.<\/span><span class=\"n\">append<\/span><span class=\"p\">(<\/span><span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">mean<\/span><span class=\"p\">(<\/span><span class=\"n\">ces_fcnn<\/span><span class=\"p\">))<\/span>\r\n    <span class=\"n\">ce_std_fcnn_list<\/span><span class=\"o\">.<\/span><span class=\"n\">append<\/span><span class=\"p\">(<\/span><span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">std<\/span><span class=\"p\">(<\/span><span class=\"n\">ces_fcnn<\/span><span class=\"p\">))<\/span>\r\n\r\n    <span class=\"c1\"># ---- NBMF ----<\/span>\r\n    <span class=\"n\">accs_nbmf<\/span><span class=\"p\">,<\/span> <span class=\"n\">ces_nbmf<\/span> <span class=\"o\">=<\/span> <span class=\"p\">[],<\/span> <span class=\"p\">[]<\/span>\r\n    <span class=\"k\">for<\/span> <span class=\"n\">s<\/span> <span class=\"ow\">in<\/span> <span class=\"nb\">range<\/span><span class=\"p\">(<\/span><span class=\"n\">num_repeats<\/span><span class=\"p\">):<\/span>\r\n        <span class=\"nb\">print<\/span><span class=\"p\">(<\/span><span class=\"sa\">f<\/span><span class=\"s2\">\"  --- NBMF \u5b9f\u9a13 <\/span><span class=\"si\">{<\/span><span class=\"n\">s<\/span><span class=\"o\">+<\/span><span class=\"mi\">1<\/span><span class=\"si\">}<\/span><span class=\"s2\">\/3 ---\"<\/span><span class=\"p\">)<\/span>\r\n        <span class=\"n\">acc_nbmf<\/span><span class=\"p\">,<\/span> <span class=\"n\">ce_nbmf<\/span><span class=\"p\">,<\/span> <span class=\"n\">W_final<\/span><span class=\"p\">,<\/span> <span class=\"n\">H_final<\/span><span class=\"p\">,<\/span> <span class=\"n\">y_pred<\/span><span class=\"p\">,<\/span> <span class=\"n\">H_test<\/span> <span class=\"o\">=<\/span> <span class=\"n\">train_nbmf_once<\/span><span class=\"p\">(<\/span>\r\n            <span class=\"n\">X_train<\/span><span class=\"p\">,<\/span> <span class=\"n\">y_train<\/span><span class=\"p\">,<\/span> <span class=\"n\">X_test<\/span><span class=\"p\">,<\/span> <span class=\"n\">y_test<\/span><span class=\"p\">,<\/span>\r\n            <span class=\"n\">k<\/span><span class=\"o\">=<\/span><span class=\"n\">k<\/span><span class=\"p\">,<\/span> <span class=\"n\">epochs<\/span><span class=\"o\">=<\/span><span class=\"n\">epochs<\/span><span class=\"p\">,<\/span> <span class=\"n\">alpha<\/span><span class=\"o\">=<\/span><span class=\"n\">alpha<\/span><span class=\"p\">,<\/span> <span class=\"n\">lr_W<\/span><span class=\"o\">=<\/span><span class=\"n\">lr_W<\/span><span class=\"p\">,<\/span> <span class=\"n\">g<\/span><span class=\"o\">=<\/span><span class=\"n\">g<\/span><span class=\"p\">,<\/span><span class=\"n\">l1<\/span><span class=\"o\">=<\/span><span class=\"n\">l1<\/span><span class=\"p\">,<\/span>\r\n            <span class=\"n\">seed<\/span><span class=\"o\">=<\/span><span class=\"n\">s<\/span><span class=\"p\">,<\/span> <span class=\"n\">num_reads<\/span><span class=\"o\">=<\/span><span class=\"mi\">50<\/span><span class=\"p\">,<\/span> <span class=\"n\">verbose<\/span><span class=\"o\">=<\/span><span class=\"kc\">False<\/span><span class=\"p\">,<\/span> <span class=\"n\">sampler<\/span><span class=\"o\">=<\/span><span class=\"n\">sampler<\/span>\r\n        <span class=\"p\">)<\/span>\r\n        <span class=\"n\">accs_nbmf<\/span><span class=\"o\">.<\/span><span class=\"n\">append<\/span><span class=\"p\">(<\/span><span class=\"n\">acc_nbmf<\/span><span class=\"p\">[<\/span><span class=\"o\">-<\/span><span class=\"mi\">1<\/span><span class=\"p\">])<\/span>\r\n        <span class=\"n\">ces_nbmf<\/span><span class=\"o\">.<\/span><span class=\"n\">append<\/span><span class=\"p\">(<\/span><span class=\"n\">ce_nbmf<\/span><span class=\"p\">[<\/span><span class=\"o\">-<\/span><span class=\"mi\">1<\/span><span class=\"p\">])<\/span>\r\n\r\n    <span class=\"n\">acc_mean_nbmf_list<\/span><span class=\"o\">.<\/span><span class=\"n\">append<\/span><span class=\"p\">(<\/span><span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">mean<\/span><span class=\"p\">(<\/span><span class=\"n\">accs_nbmf<\/span><span class=\"p\">))<\/span>\r\n    <span class=\"n\">acc_std_nbmf_list<\/span><span class=\"o\">.<\/span><span class=\"n\">append<\/span><span class=\"p\">(<\/span><span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">std<\/span><span class=\"p\">(<\/span><span class=\"n\">accs_nbmf<\/span><span class=\"p\">))<\/span>\r\n    <span class=\"n\">ce_mean_nbmf_list<\/span><span class=\"o\">.<\/span><span class=\"n\">append<\/span><span class=\"p\">(<\/span><span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">mean<\/span><span class=\"p\">(<\/span><span class=\"n\">ces_nbmf<\/span><span class=\"p\">))<\/span>\r\n    <span class=\"n\">ce_std_nbmf_list<\/span><span class=\"o\">.<\/span><span class=\"n\">append<\/span><span class=\"p\">(<\/span><span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">std<\/span><span class=\"p\">(<\/span><span class=\"n\">ces_nbmf<\/span><span class=\"p\">))<\/span>\r\n\r\n<span class=\"n\">acc_mean_fcnn<\/span> <span class=\"o\">=<\/span> <span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">array<\/span><span class=\"p\">(<\/span><span class=\"n\">acc_mean_fcnn_list<\/span><span class=\"p\">)<\/span>\r\n<span class=\"n\">acc_std_fcnn<\/span> <span class=\"o\">=<\/span> <span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">array<\/span><span class=\"p\">(<\/span><span class=\"n\">acc_std_fcnn_list<\/span><span class=\"p\">)<\/span>\r\n<span class=\"n\">ce_mean_fcnn<\/span> <span class=\"o\">=<\/span> <span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">array<\/span><span class=\"p\">(<\/span><span class=\"n\">ce_mean_fcnn_list<\/span><span class=\"p\">)<\/span>\r\n<span class=\"n\">ce_std_fcnn<\/span> <span class=\"o\">=<\/span> <span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">array<\/span><span class=\"p\">(<\/span><span class=\"n\">ce_std_fcnn_list<\/span><span class=\"p\">)<\/span>\r\n<span class=\"n\">acc_mean_nbmf<\/span> <span class=\"o\">=<\/span> <span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">array<\/span><span class=\"p\">(<\/span><span class=\"n\">acc_mean_nbmf_list<\/span><span class=\"p\">)<\/span>\r\n<span class=\"n\">acc_std_nbmf<\/span> <span class=\"o\">=<\/span> <span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">array<\/span><span class=\"p\">(<\/span><span class=\"n\">acc_std_nbmf_list<\/span><span class=\"p\">)<\/span>\r\n<span class=\"n\">ce_mean_nbmf<\/span> <span class=\"o\">=<\/span> <span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">array<\/span><span class=\"p\">(<\/span><span class=\"n\">ce_mean_nbmf_list<\/span><span class=\"p\">)<\/span>\r\n<span class=\"n\">ce_std_nbmf<\/span> <span class=\"o\">=<\/span> <span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">array<\/span><span class=\"p\">(<\/span><span class=\"n\">ce_std_nbmf_list<\/span><span class=\"p\">)<\/span>\r\n<\/pre>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<details>\n<summary style=\"cursor: pointer; font-weight: bold; padding: 10px; background-color: #f0f0f0; border-radius: 5px;\">\n\u5b9f\u9a13\u7d50\u679c\u3092\u8868\u793a\u3059\u308b\uff08\u3053\u3053\u3092\u30af\u30ea\u30c3\u30af\uff09<br \/>\n<\/summary>\n<div class=\"output_wrapper\">\n<div class=\"output\">\n<div class=\"output_area\">\n<div class=\"prompt\"><\/div>\n<div class=\"output_subarea output_stream output_stdout output_text\">\n<pre>--- m_train = 50 ---\r\n  --- NBMF \u5b9f\u9a13 1\/3 ---\r\n[Epoch 01] TestAcc=30.60%  CE=2.1981\r\n[Epoch 02] TestAcc=62.20%  CE=1.6639\r\n[Epoch 03] TestAcc=63.80%  CE=1.5467\r\n[Epoch 04] TestAcc=64.80%  CE=1.4787\r\n[Epoch 05] TestAcc=65.40%  CE=1.4792\r\n[Epoch 06] TestAcc=65.60%  CE=1.4804\r\n[Epoch 07] TestAcc=65.60%  CE=1.4755\r\n[Epoch 08] TestAcc=66.20%  CE=1.4775\r\n[Epoch 09] TestAcc=64.80%  CE=1.4750\r\n[Epoch 10] TestAcc=64.60%  CE=1.4787\r\n  --- NBMF \u5b9f\u9a13 2\/3 ---\r\n[Epoch 01] TestAcc=35.60%  CE=2.1371\r\n[Epoch 02] TestAcc=66.40%  CE=1.4351\r\n[Epoch 03] TestAcc=67.40%  CE=1.4289\r\n[Epoch 04] TestAcc=66.80%  CE=1.4095\r\n[Epoch 05] TestAcc=67.20%  CE=1.4106\r\n[Epoch 06] TestAcc=66.60%  CE=1.4138\r\n[Epoch 07] TestAcc=67.80%  CE=1.4147\r\n[Epoch 08] TestAcc=67.40%  CE=1.4090\r\n[Epoch 09] TestAcc=66.80%  CE=1.4046\r\n[Epoch 10] TestAcc=67.20%  CE=1.4052\r\n  --- NBMF \u5b9f\u9a13 3\/3 ---\r\n[Epoch 01] TestAcc=36.80%  CE=2.1974\r\n[Epoch 02] TestAcc=63.20%  CE=1.5463\r\n[Epoch 03] TestAcc=62.40%  CE=1.5454\r\n[Epoch 04] TestAcc=62.00%  CE=1.5447\r\n[Epoch 05] TestAcc=62.40%  CE=1.5434\r\n[Epoch 06] TestAcc=62.20%  CE=1.5449\r\n[Epoch 07] TestAcc=60.80%  CE=1.5438\r\n[Epoch 08] TestAcc=62.00%  CE=1.5480\r\n[Epoch 09] TestAcc=62.40%  CE=1.5405\r\n[Epoch 10] TestAcc=60.80%  CE=1.5471\r\n\r\n--- m_train = 75 ---\r\n  --- NBMF \u5b9f\u9a13 1\/3 ---\r\n[Epoch 01] TestAcc=31.00%  CE=2.2189\r\n[Epoch 02] TestAcc=59.60%  CE=1.8389\r\n[Epoch 03] TestAcc=67.00%  CE=1.3627\r\n[Epoch 04] TestAcc=66.40%  CE=1.3241\r\n[Epoch 05] TestAcc=66.00%  CE=1.3162\r\n[Epoch 06] TestAcc=65.40%  CE=1.3154\r\n[Epoch 07] TestAcc=65.40%  CE=1.3264\r\n[Epoch 08] TestAcc=64.40%  CE=1.3222\r\n[Epoch 09] TestAcc=65.00%  CE=1.3146\r\n[Epoch 10] TestAcc=65.00%  CE=1.3117\r\n  --- NBMF \u5b9f\u9a13 2\/3 ---\r\n[Epoch 01] TestAcc=36.80%  CE=2.1676\r\n[Epoch 02] TestAcc=47.20%  CE=2.0624\r\n[Epoch 03] TestAcc=58.40%  CE=1.9104\r\n[Epoch 04] TestAcc=62.60%  CE=1.4387\r\n[Epoch 05] TestAcc=64.40%  CE=1.3577\r\n[Epoch 06] TestAcc=64.00%  CE=1.3506\r\n[Epoch 07] TestAcc=64.00%  CE=1.3447\r\n[Epoch 08] TestAcc=64.40%  CE=1.3414\r\n[Epoch 09] TestAcc=64.60%  CE=1.3401\r\n[Epoch 10] TestAcc=65.00%  CE=1.3404\r\n  --- NBMF \u5b9f\u9a13 3\/3 ---\r\n[Epoch 01] TestAcc=31.00%  CE=2.2016\r\n[Epoch 02] TestAcc=45.20%  CE=1.9640\r\n[Epoch 03] TestAcc=49.80%  CE=1.8815\r\n[Epoch 04] TestAcc=50.60%  CE=1.8822\r\n[Epoch 05] TestAcc=55.40%  CE=1.7821\r\n[Epoch 06] TestAcc=63.80%  CE=1.3953\r\n[Epoch 07] TestAcc=65.00%  CE=1.3531\r\n[Epoch 08] TestAcc=65.00%  CE=1.3536\r\n[Epoch 09] TestAcc=64.00%  CE=1.3586\r\n[Epoch 10] TestAcc=64.60%  CE=1.3580\r\n\r\n--- m_train = 100 ---\r\n  --- NBMF \u5b9f\u9a13 1\/3 ---\r\n[Epoch 01] TestAcc=33.20%  CE=2.1985\r\n[Epoch 02] TestAcc=63.00%  CE=1.5476\r\n[Epoch 03] TestAcc=63.80%  CE=1.3889\r\n[Epoch 04] TestAcc=63.80%  CE=1.3941\r\n[Epoch 05] TestAcc=64.20%  CE=1.3929\r\n[Epoch 06] TestAcc=64.20%  CE=1.3856\r\n[Epoch 07] TestAcc=63.40%  CE=1.3927\r\n[Epoch 08] TestAcc=64.00%  CE=1.3973\r\n[Epoch 09] TestAcc=64.00%  CE=1.3860\r\n[Epoch 10] TestAcc=64.40%  CE=1.3886\r\n  --- NBMF \u5b9f\u9a13 2\/3 ---\r\n[Epoch 01] TestAcc=32.00%  CE=2.2067\r\n[Epoch 02] TestAcc=59.80%  CE=1.5729\r\n[Epoch 03] TestAcc=62.80%  CE=1.4103\r\n[Epoch 04] TestAcc=62.20%  CE=1.3834\r\n[Epoch 05] TestAcc=61.80%  CE=1.3815\r\n[Epoch 06] TestAcc=60.60%  CE=1.3817\r\n[Epoch 07] TestAcc=61.00%  CE=1.3774\r\n[Epoch 08] TestAcc=62.00%  CE=1.3664\r\n[Epoch 09] TestAcc=61.40%  CE=1.3674\r\n[Epoch 10] TestAcc=61.40%  CE=1.3765\r\n  --- NBMF \u5b9f\u9a13 3\/3 ---\r\n[Epoch 01] TestAcc=36.40%  CE=2.1776\r\n[Epoch 02] TestAcc=62.40%  CE=1.5959\r\n[Epoch 03] TestAcc=63.80%  CE=1.4232\r\n[Epoch 04] TestAcc=65.60%  CE=1.3603\r\n[Epoch 05] TestAcc=64.80%  CE=1.3754\r\n[Epoch 06] TestAcc=65.00%  CE=1.3730\r\n[Epoch 07] TestAcc=65.20%  CE=1.3589\r\n[Epoch 08] TestAcc=65.00%  CE=1.3724\r\n[Epoch 09] TestAcc=65.00%  CE=1.3631\r\n[Epoch 10] TestAcc=64.20%  CE=1.3755\r\n\r\n--- m_train = 125 ---\r\n  --- NBMF \u5b9f\u9a13 1\/3 ---\r\n[Epoch 01] TestAcc=36.20%  CE=2.1944\r\n[Epoch 02] TestAcc=64.80%  CE=1.5164\r\n[Epoch 03] TestAcc=67.80%  CE=1.3391\r\n[Epoch 04] TestAcc=68.80%  CE=1.3287\r\n[Epoch 05] TestAcc=67.80%  CE=1.3362\r\n[Epoch 06] TestAcc=67.60%  CE=1.3297\r\n[Epoch 07] TestAcc=67.80%  CE=1.3331\r\n[Epoch 08] TestAcc=68.60%  CE=1.3321\r\n[Epoch 09] TestAcc=68.20%  CE=1.3283\r\n[Epoch 10] TestAcc=67.80%  CE=1.3333\r\n  --- NBMF \u5b9f\u9a13 2\/3 ---\r\n[Epoch 01] TestAcc=42.00%  CE=2.1226\r\n[Epoch 02] TestAcc=53.00%  CE=1.9646\r\n[Epoch 03] TestAcc=64.80%  CE=1.4812\r\n[Epoch 04] TestAcc=68.40%  CE=1.3459\r\n[Epoch 05] TestAcc=67.00%  CE=1.3533\r\n[Epoch 06] TestAcc=68.00%  CE=1.3483\r\n[Epoch 07] TestAcc=68.00%  CE=1.3478\r\n[Epoch 08] TestAcc=66.80%  CE=1.3529\r\n[Epoch 09] TestAcc=66.60%  CE=1.3572\r\n[Epoch 10] TestAcc=68.00%  CE=1.3546\r\n  --- NBMF \u5b9f\u9a13 3\/3 ---\r\n[Epoch 01] TestAcc=31.80%  CE=2.1889\r\n[Epoch 02] TestAcc=65.00%  CE=1.5568\r\n[Epoch 03] TestAcc=67.80%  CE=1.3654\r\n[Epoch 04] TestAcc=69.00%  CE=1.3172\r\n[Epoch 05] TestAcc=70.40%  CE=1.2922\r\n[Epoch 06] TestAcc=70.00%  CE=1.2821\r\n[Epoch 07] TestAcc=69.60%  CE=1.2827\r\n[Epoch 08] TestAcc=69.40%  CE=1.2839\r\n[Epoch 09] TestAcc=70.60%  CE=1.2825\r\n[Epoch 10] TestAcc=69.00%  CE=1.2967\r\n\r\n--- m_train = 150 ---\r\n  --- NBMF \u5b9f\u9a13 1\/3 ---\r\n[Epoch 01] TestAcc=46.00%  CE=2.0570\r\n[Epoch 02] TestAcc=66.60%  CE=1.5906\r\n[Epoch 03] TestAcc=65.40%  CE=1.4393\r\n[Epoch 04] TestAcc=66.40%  CE=1.3598\r\n[Epoch 05] TestAcc=65.40%  CE=1.3386\r\n[Epoch 06] TestAcc=65.60%  CE=1.3359\r\n[Epoch 07] TestAcc=66.40%  CE=1.3391\r\n[Epoch 08] TestAcc=65.20%  CE=1.3339\r\n[Epoch 09] TestAcc=65.20%  CE=1.3347\r\n[Epoch 10] TestAcc=66.40%  CE=1.3349\r\n  --- NBMF \u5b9f\u9a13 2\/3 ---\r\n[Epoch 01] TestAcc=31.20%  CE=2.2064\r\n[Epoch 02] TestAcc=67.00%  CE=1.4999\r\n[Epoch 03] TestAcc=71.00%  CE=1.2874\r\n[Epoch 04] TestAcc=70.40%  CE=1.2398\r\n[Epoch 05] TestAcc=70.80%  CE=1.2448\r\n[Epoch 06] TestAcc=70.80%  CE=1.2457\r\n[Epoch 07] TestAcc=71.60%  CE=1.2381\r\n[Epoch 08] TestAcc=70.60%  CE=1.2447\r\n[Epoch 09] TestAcc=71.20%  CE=1.2411\r\n[Epoch 10] TestAcc=71.00%  CE=1.2375\r\n  --- NBMF \u5b9f\u9a13 3\/3 ---\r\n[Epoch 01] TestAcc=28.00%  CE=2.2173\r\n[Epoch 02] TestAcc=66.00%  CE=1.4960\r\n[Epoch 03] TestAcc=68.40%  CE=1.2950\r\n[Epoch 04] TestAcc=67.60%  CE=1.2792\r\n[Epoch 05] TestAcc=67.20%  CE=1.2784\r\n[Epoch 06] TestAcc=67.40%  CE=1.2725\r\n[Epoch 07] TestAcc=68.20%  CE=1.2613\r\n[Epoch 08] TestAcc=69.40%  CE=1.2588\r\n[Epoch 09] TestAcc=67.80%  CE=1.2610\r\n[Epoch 10] TestAcc=68.60%  CE=1.2601\r\n\r\n--- m_train = 175 ---\r\n  --- NBMF \u5b9f\u9a13 1\/3 ---\r\n[Epoch 01] TestAcc=32.80%  CE=2.1954\r\n[Epoch 02] TestAcc=63.60%  CE=1.6437\r\n[Epoch 03] TestAcc=66.60%  CE=1.4494\r\n[Epoch 04] TestAcc=68.00%  CE=1.3474\r\n[Epoch 05] TestAcc=68.60%  CE=1.3418\r\n[Epoch 06] TestAcc=68.80%  CE=1.3114\r\n[Epoch 07] TestAcc=68.20%  CE=1.3049\r\n[Epoch 08] TestAcc=68.40%  CE=1.2870\r\n[Epoch 09] TestAcc=68.80%  CE=1.2853\r\n[Epoch 10] TestAcc=67.60%  CE=1.2901\r\n  --- NBMF \u5b9f\u9a13 2\/3 ---\r\n[Epoch 01] TestAcc=29.60%  CE=2.2013\r\n[Epoch 02] TestAcc=63.00%  CE=1.6053\r\n[Epoch 03] TestAcc=61.60%  CE=1.4201\r\n[Epoch 04] TestAcc=65.40%  CE=1.3777\r\n[Epoch 05] TestAcc=65.20%  CE=1.3760\r\n[Epoch 06] TestAcc=65.80%  CE=1.3825\r\n[Epoch 07] TestAcc=65.20%  CE=1.3707\r\n[Epoch 08] TestAcc=64.80%  CE=1.3766\r\n[Epoch 09] TestAcc=66.00%  CE=1.3702\r\n[Epoch 10] TestAcc=64.80%  CE=1.3827\r\n  --- NBMF \u5b9f\u9a13 3\/3 ---\r\n[Epoch 01] TestAcc=31.60%  CE=2.2300\r\n[Epoch 02] TestAcc=62.80%  CE=1.6236\r\n[Epoch 03] TestAcc=66.00%  CE=1.4081\r\n[Epoch 04] TestAcc=65.20%  CE=1.3999\r\n[Epoch 05] TestAcc=65.40%  CE=1.3938\r\n[Epoch 06] TestAcc=65.80%  CE=1.3953\r\n[Epoch 07] TestAcc=64.80%  CE=1.4006\r\n[Epoch 08] TestAcc=65.20%  CE=1.4001\r\n[Epoch 09] TestAcc=66.00%  CE=1.3897\r\n[Epoch 10] TestAcc=65.80%  CE=1.3849\r\n\r\n--- m_train = 200 ---\r\n  --- NBMF \u5b9f\u9a13 1\/3 ---\r\n[Epoch 01] TestAcc=30.40%  CE=2.2203\r\n[Epoch 02] TestAcc=52.40%  CE=1.9877\r\n[Epoch 03] TestAcc=60.80%  CE=1.8334\r\n[Epoch 04] TestAcc=70.20%  CE=1.4161\r\n[Epoch 05] TestAcc=70.40%  CE=1.3154\r\n[Epoch 06] TestAcc=70.80%  CE=1.2742\r\n[Epoch 07] TestAcc=69.60%  CE=1.2764\r\n[Epoch 08] TestAcc=69.00%  CE=1.2559\r\n[Epoch 09] TestAcc=70.20%  CE=1.2539\r\n[Epoch 10] TestAcc=69.00%  CE=1.2553\r\n  --- NBMF \u5b9f\u9a13 2\/3 ---\r\n[Epoch 01] TestAcc=28.00%  CE=2.2333\r\n[Epoch 02] TestAcc=63.60%  CE=1.5435\r\n[Epoch 03] TestAcc=66.80%  CE=1.3359\r\n[Epoch 04] TestAcc=68.80%  CE=1.2621\r\n[Epoch 05] TestAcc=68.80%  CE=1.2601\r\n[Epoch 06] TestAcc=67.40%  CE=1.2742\r\n[Epoch 07] TestAcc=69.00%  CE=1.2650\r\n[Epoch 08] TestAcc=69.00%  CE=1.2700\r\n[Epoch 09] TestAcc=69.40%  CE=1.2596\r\n[Epoch 10] TestAcc=69.00%  CE=1.2605\r\n  --- NBMF \u5b9f\u9a13 3\/3 ---\r\n[Epoch 01] TestAcc=40.60%  CE=2.1574\r\n[Epoch 02] TestAcc=65.40%  CE=1.5926\r\n[Epoch 03] TestAcc=68.20%  CE=1.3823\r\n[Epoch 04] TestAcc=68.80%  CE=1.3040\r\n[Epoch 05] TestAcc=67.20%  CE=1.2893\r\n[Epoch 06] TestAcc=68.40%  CE=1.2790\r\n[Epoch 07] TestAcc=69.00%  CE=1.2797\r\n[Epoch 08] TestAcc=68.60%  CE=1.2834\r\n[Epoch 09] TestAcc=67.80%  CE=1.2768\r\n[Epoch 10] TestAcc=67.60%  CE=1.2813\r\n\r\n--- m_train = 225 ---\r\n  --- NBMF \u5b9f\u9a13 1\/3 ---\r\n[Epoch 01] TestAcc=25.20%  CE=2.2348\r\n[Epoch 02] TestAcc=68.80%  CE=1.5863\r\n[Epoch 03] TestAcc=71.20%  CE=1.3270\r\n[Epoch 04] TestAcc=69.20%  CE=1.2585\r\n[Epoch 05] TestAcc=70.80%  CE=1.2152\r\n[Epoch 06] TestAcc=72.20%  CE=1.1737\r\n[Epoch 07] TestAcc=72.00%  CE=1.1663\r\n[Epoch 08] TestAcc=72.20%  CE=1.1590\r\n[Epoch 09] TestAcc=72.00%  CE=1.1593\r\n[Epoch 10] TestAcc=73.60%  CE=1.1525\r\n  --- NBMF \u5b9f\u9a13 2\/3 ---\r\n[Epoch 01] TestAcc=28.60%  CE=2.2174\r\n[Epoch 02] TestAcc=65.80%  CE=1.5795\r\n[Epoch 03] TestAcc=69.00%  CE=1.3761\r\n[Epoch 04] TestAcc=67.40%  CE=1.3415\r\n[Epoch 05] TestAcc=68.00%  CE=1.2994\r\n[Epoch 06] TestAcc=68.40%  CE=1.2310\r\n[Epoch 07] TestAcc=68.40%  CE=1.2399\r\n[Epoch 08] TestAcc=67.80%  CE=1.2314\r\n[Epoch 09] TestAcc=67.40%  CE=1.2310\r\n[Epoch 10] TestAcc=68.20%  CE=1.2284\r\n  --- NBMF \u5b9f\u9a13 3\/3 ---\r\n[Epoch 01] TestAcc=29.00%  CE=2.2206\r\n[Epoch 02] TestAcc=70.40%  CE=1.5858\r\n[Epoch 03] TestAcc=71.40%  CE=1.3398\r\n[Epoch 04] TestAcc=71.20%  CE=1.2617\r\n[Epoch 05] TestAcc=70.60%  CE=1.2671\r\n[Epoch 06] TestAcc=69.60%  CE=1.2724\r\n[Epoch 07] TestAcc=71.00%  CE=1.2654\r\n[Epoch 08] TestAcc=70.40%  CE=1.2632\r\n[Epoch 09] TestAcc=70.60%  CE=1.2688\r\n[Epoch 10] TestAcc=70.20%  CE=1.2661\r\n\r\n--- m_train = 250 ---\r\n  --- NBMF \u5b9f\u9a13 1\/3 ---\r\n[Epoch 01] TestAcc=35.60%  CE=2.1945\r\n[Epoch 02] TestAcc=66.20%  CE=1.6137\r\n[Epoch 03] TestAcc=71.80%  CE=1.3388\r\n[Epoch 04] TestAcc=72.00%  CE=1.2302\r\n[Epoch 05] TestAcc=73.20%  CE=1.2253\r\n[Epoch 06] TestAcc=72.80%  CE=1.2238\r\n[Epoch 07] TestAcc=73.40%  CE=1.2242\r\n[Epoch 08] TestAcc=73.00%  CE=1.2224\r\n[Epoch 09] TestAcc=72.80%  CE=1.2231\r\n[Epoch 10] TestAcc=71.80%  CE=1.2290\r\n  --- NBMF \u5b9f\u9a13 2\/3 ---\r\n[Epoch 01] TestAcc=44.40%  CE=2.1238\r\n[Epoch 02] TestAcc=72.20%  CE=1.5569\r\n[Epoch 03] TestAcc=72.80%  CE=1.3053\r\n[Epoch 04] TestAcc=74.00%  CE=1.3109\r\n[Epoch 05] TestAcc=74.20%  CE=1.3078\r\n[Epoch 06] TestAcc=74.00%  CE=1.2145\r\n[Epoch 07] TestAcc=73.40%  CE=1.2107\r\n[Epoch 08] TestAcc=73.20%  CE=1.1671\r\n[Epoch 09] TestAcc=74.40%  CE=1.1282\r\n[Epoch 10] TestAcc=73.60%  CE=1.1263\r\n  --- NBMF \u5b9f\u9a13 3\/3 ---\r\n[Epoch 01] TestAcc=36.20%  CE=2.1708\r\n[Epoch 02] TestAcc=68.60%  CE=1.5747\r\n[Epoch 03] TestAcc=69.60%  CE=1.3108\r\n[Epoch 04] TestAcc=71.60%  CE=1.2575\r\n[Epoch 05] TestAcc=69.40%  CE=1.1764\r\n[Epoch 06] TestAcc=69.40%  CE=1.1802\r\n[Epoch 07] TestAcc=70.00%  CE=1.1729\r\n[Epoch 08] TestAcc=69.80%  CE=1.1727\r\n[Epoch 09] TestAcc=69.60%  CE=1.1720\r\n[Epoch 10] TestAcc=70.00%  CE=1.1806\r\n\r\n--- m_train = 275 ---\r\n  --- NBMF \u5b9f\u9a13 1\/3 ---\r\n[Epoch 01] TestAcc=32.60%  CE=2.2223\r\n[Epoch 02] TestAcc=62.80%  CE=1.6247\r\n[Epoch 03] TestAcc=66.00%  CE=1.4030\r\n[Epoch 04] TestAcc=67.00%  CE=1.3197\r\n[Epoch 05] TestAcc=68.60%  CE=1.2703\r\n[Epoch 06] TestAcc=71.20%  CE=1.1632\r\n[Epoch 07] TestAcc=71.60%  CE=1.1624\r\n[Epoch 08] TestAcc=72.20%  CE=1.1522\r\n[Epoch 09] TestAcc=71.60%  CE=1.1645\r\n[Epoch 10] TestAcc=71.20%  CE=1.1645\r\n  --- NBMF \u5b9f\u9a13 2\/3 ---\r\n[Epoch 01] TestAcc=26.00%  CE=2.2367\r\n[Epoch 02] TestAcc=69.20%  CE=1.5122\r\n[Epoch 03] TestAcc=70.80%  CE=1.2938\r\n[Epoch 04] TestAcc=74.20%  CE=1.1944\r\n[Epoch 05] TestAcc=72.80%  CE=1.1722\r\n[Epoch 06] TestAcc=72.80%  CE=1.1218\r\n[Epoch 07] TestAcc=73.00%  CE=1.1253\r\n[Epoch 08] TestAcc=73.40%  CE=1.1224\r\n[Epoch 09] TestAcc=74.00%  CE=1.1198\r\n[Epoch 10] TestAcc=73.60%  CE=1.1124\r\n  --- NBMF \u5b9f\u9a13 3\/3 ---\r\n[Epoch 01] TestAcc=29.80%  CE=2.2335\r\n[Epoch 02] TestAcc=68.20%  CE=1.6066\r\n[Epoch 03] TestAcc=70.20%  CE=1.3788\r\n[Epoch 04] TestAcc=70.40%  CE=1.3250\r\n[Epoch 05] TestAcc=70.20%  CE=1.3106\r\n[Epoch 06] TestAcc=73.00%  CE=1.1949\r\n[Epoch 07] TestAcc=73.40%  CE=1.1917\r\n[Epoch 08] TestAcc=73.40%  CE=1.1829\r\n[Epoch 09] TestAcc=73.20%  CE=1.1804\r\n[Epoch 10] TestAcc=73.40%  CE=1.1791\r\n\r\n--- m_train = 300 ---\r\n  --- NBMF \u5b9f\u9a13 1\/3 ---\r\n[Epoch 01] TestAcc=17.00%  CE=2.2672\r\n[Epoch 02] TestAcc=65.00%  CE=1.5885\r\n[Epoch 03] TestAcc=68.80%  CE=1.3041\r\n[Epoch 04] TestAcc=72.40%  CE=1.1717\r\n[Epoch 05] TestAcc=73.80%  CE=1.1727\r\n[Epoch 06] TestAcc=72.60%  CE=1.1764\r\n[Epoch 07] TestAcc=73.40%  CE=1.1676\r\n[Epoch 08] TestAcc=73.00%  CE=1.1834\r\n[Epoch 09] TestAcc=72.80%  CE=1.1756\r\n[Epoch 10] TestAcc=73.60%  CE=1.1699\r\n  --- NBMF \u5b9f\u9a13 2\/3 ---\r\n[Epoch 01] TestAcc=32.60%  CE=2.2074\r\n[Epoch 02] TestAcc=68.80%  CE=1.5622\r\n[Epoch 03] TestAcc=67.80%  CE=1.3007\r\n[Epoch 04] TestAcc=70.00%  CE=1.1990\r\n[Epoch 05] TestAcc=70.80%  CE=1.2027\r\n[Epoch 06] TestAcc=71.00%  CE=1.1985\r\n[Epoch 07] TestAcc=71.60%  CE=1.1947\r\n[Epoch 08] TestAcc=70.00%  CE=1.2094\r\n[Epoch 09] TestAcc=69.80%  CE=1.2000\r\n[Epoch 10] TestAcc=70.40%  CE=1.2051\r\n  --- NBMF \u5b9f\u9a13 3\/3 ---\r\n[Epoch 01] TestAcc=36.20%  CE=2.1899\r\n[Epoch 02] TestAcc=58.60%  CE=1.9833\r\n[Epoch 03] TestAcc=71.80%  CE=1.4741\r\n[Epoch 04] TestAcc=72.60%  CE=1.2910\r\n[Epoch 05] TestAcc=72.60%  CE=1.2710\r\n[Epoch 06] TestAcc=72.60%  CE=1.2693\r\n[Epoch 07] TestAcc=72.80%  CE=1.2748\r\n[Epoch 08] TestAcc=72.40%  CE=1.2765\r\n[Epoch 09] TestAcc=72.40%  CE=1.2755\r\n[Epoch 10] TestAcc=73.20%  CE=1.2752\r\n\r\n--- m_train = 325 ---\r\n  --- NBMF \u5b9f\u9a13 1\/3 ---\r\n[Epoch 01] TestAcc=24.00%  CE=2.2571\r\n[Epoch 02] TestAcc=47.40%  CE=2.0823\r\n[Epoch 03] TestAcc=66.60%  CE=1.5272\r\n[Epoch 04] TestAcc=69.00%  CE=1.3704\r\n[Epoch 05] TestAcc=67.80%  CE=1.3754\r\n[Epoch 06] TestAcc=67.80%  CE=1.3754\r\n[Epoch 07] TestAcc=68.40%  CE=1.3675\r\n[Epoch 08] TestAcc=69.80%  CE=1.3651\r\n[Epoch 09] TestAcc=67.00%  CE=1.3770\r\n[Epoch 10] TestAcc=68.20%  CE=1.3779\r\n  --- NBMF \u5b9f\u9a13 2\/3 ---\r\n[Epoch 01] TestAcc=28.20%  CE=2.2278\r\n[Epoch 02] TestAcc=70.40%  CE=1.5695\r\n[Epoch 03] TestAcc=70.80%  CE=1.3178\r\n[Epoch 04] TestAcc=70.00%  CE=1.1955\r\n[Epoch 05] TestAcc=71.00%  CE=1.1662\r\n[Epoch 06] TestAcc=72.00%  CE=1.1447\r\n[Epoch 07] TestAcc=72.60%  CE=1.1421\r\n[Epoch 08] TestAcc=71.60%  CE=1.1510\r\n[Epoch 09] TestAcc=71.40%  CE=1.1436\r\n[Epoch 10] TestAcc=72.00%  CE=1.1520\r\n  --- NBMF \u5b9f\u9a13 3\/3 ---\r\n[Epoch 01] TestAcc=30.00%  CE=2.1954\r\n[Epoch 02] TestAcc=66.60%  CE=1.5897\r\n[Epoch 03] TestAcc=67.20%  CE=1.3502\r\n[Epoch 04] TestAcc=68.80%  CE=1.2314\r\n[Epoch 05] TestAcc=68.80%  CE=1.2364\r\n[Epoch 06] TestAcc=71.40%  CE=1.1558\r\n[Epoch 07] TestAcc=72.40%  CE=1.1590\r\n[Epoch 08] TestAcc=71.60%  CE=1.1649\r\n[Epoch 09] TestAcc=71.40%  CE=1.1639\r\n[Epoch 10] TestAcc=71.40%  CE=1.1555\r\n\r\n--- m_train = 350 ---\r\n  --- NBMF \u5b9f\u9a13 1\/3 ---\r\n[Epoch 01] TestAcc=30.80%  CE=2.2281\r\n[Epoch 02] TestAcc=68.20%  CE=1.5876\r\n[Epoch 03] TestAcc=70.20%  CE=1.3455\r\n[Epoch 04] TestAcc=71.40%  CE=1.2837\r\n[Epoch 05] TestAcc=71.20%  CE=1.2388\r\n[Epoch 06] TestAcc=71.00%  CE=1.2339\r\n[Epoch 07] TestAcc=71.00%  CE=1.2449\r\n[Epoch 08] TestAcc=70.80%  CE=1.2338\r\n[Epoch 09] TestAcc=71.00%  CE=1.2388\r\n[Epoch 10] TestAcc=71.00%  CE=1.2414\r\n  --- NBMF \u5b9f\u9a13 2\/3 ---\r\n[Epoch 01] TestAcc=43.40%  CE=2.1387\r\n[Epoch 02] TestAcc=64.20%  CE=1.5753\r\n[Epoch 03] TestAcc=69.20%  CE=1.3721\r\n[Epoch 04] TestAcc=69.60%  CE=1.2728\r\n[Epoch 05] TestAcc=69.20%  CE=1.2059\r\n[Epoch 06] TestAcc=68.80%  CE=1.2131\r\n[Epoch 07] TestAcc=69.00%  CE=1.2194\r\n[Epoch 08] TestAcc=69.20%  CE=1.2148\r\n[Epoch 09] TestAcc=68.20%  CE=1.2115\r\n[Epoch 10] TestAcc=68.00%  CE=1.2131\r\n  --- NBMF \u5b9f\u9a13 3\/3 ---\r\n[Epoch 01] TestAcc=25.00%  CE=2.2399\r\n[Epoch 02] TestAcc=65.80%  CE=1.6277\r\n[Epoch 03] TestAcc=65.20%  CE=1.3659\r\n[Epoch 04] TestAcc=68.60%  CE=1.2847\r\n[Epoch 05] TestAcc=69.80%  CE=1.2797\r\n[Epoch 06] TestAcc=68.80%  CE=1.2791\r\n[Epoch 07] TestAcc=69.60%  CE=1.2842\r\n[Epoch 08] TestAcc=69.00%  CE=1.2814\r\n[Epoch 09] TestAcc=69.60%  CE=1.2832\r\n[Epoch 10] TestAcc=70.00%  CE=1.2844\r\n<\/pre>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<\/details>\n<div class=\"cell border-box-sizing text_cell rendered\">\n<div class=\"inner_cell\">\n<div class=\"text_cell_render border-box-sizing rendered_html\">\n<p>\u7d50\u679c\u3092\u8868\u793a\u3057\u307e\u3059\u3002<\/p>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing code_cell rendered\">\n<div class=\"input\">\n<div class=\"inner_cell\">\n<div class=\"input_area\">\n<div class=\"highlight hl-python\">\n<pre><span><\/span><span class=\"nb\">print<\/span><span class=\"p\">(<\/span><span class=\"s2\">\"m_train |  FCNN_Acc(%)  NBMF_Acc(%)  |  FCNN_CE   NBMF_CE\"<\/span><span class=\"p\">)<\/span>\r\n<span class=\"k\">for<\/span> <span class=\"n\">m<\/span><span class=\"p\">,<\/span> <span class=\"n\">af<\/span><span class=\"p\">,<\/span> <span class=\"n\">an<\/span><span class=\"p\">,<\/span> <span class=\"n\">cf<\/span><span class=\"p\">,<\/span> <span class=\"n\">cn<\/span> <span class=\"ow\">in<\/span> <span class=\"nb\">zip<\/span><span class=\"p\">(<\/span><span class=\"n\">train_sizes<\/span><span class=\"p\">,<\/span> <span class=\"n\">acc_mean_fcnn<\/span><span class=\"p\">,<\/span> <span class=\"n\">acc_mean_nbmf<\/span><span class=\"p\">,<\/span> <span class=\"n\">ce_mean_fcnn<\/span><span class=\"p\">,<\/span> <span class=\"n\">ce_mean_nbmf<\/span><span class=\"p\">):<\/span>\r\n    <span class=\"nb\">print<\/span><span class=\"p\">(<\/span><span class=\"sa\">f<\/span><span class=\"s2\">\"<\/span><span class=\"si\">{<\/span><span class=\"n\">m<\/span><span class=\"si\">:<\/span><span class=\"s2\">6d<\/span><span class=\"si\">}<\/span><span class=\"s2\"> |   <\/span><span class=\"si\">{<\/span><span class=\"n\">af<\/span><span class=\"o\">*<\/span><span class=\"mi\">100<\/span><span class=\"si\">:<\/span><span class=\"s2\">7.2f<\/span><span class=\"si\">}<\/span><span class=\"s2\">      <\/span><span class=\"si\">{<\/span><span class=\"n\">an<\/span><span class=\"o\">*<\/span><span class=\"mi\">100<\/span><span class=\"si\">:<\/span><span class=\"s2\">7.2f<\/span><span class=\"si\">}<\/span><span class=\"s2\">   |  <\/span><span class=\"si\">{<\/span><span class=\"n\">cf<\/span><span class=\"si\">:<\/span><span class=\"s2\">8.4f<\/span><span class=\"si\">}<\/span><span class=\"s2\">  <\/span><span class=\"si\">{<\/span><span class=\"n\">cn<\/span><span class=\"si\">:<\/span><span class=\"s2\">8.4f<\/span><span class=\"si\">}<\/span><span class=\"s2\">\"<\/span><span class=\"p\">)<\/span>\r\n<\/pre>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"output_wrapper\">\n<div class=\"output\">\n<div class=\"output_area\">\n<div class=\"prompt\"><\/div>\n<div class=\"output_subarea output_stream output_stdout output_text\">\n<pre>m_train |  FCNN_Acc(%)  NBMF_Acc(%)  |  FCNN_CE   NBMF_CE\r\n    50 |     27.20        64.20   |    2.2426    1.4770\r\n    75 |     28.87        64.87   |    2.2279    1.3367\r\n   100 |     33.87        63.33   |    2.1930    1.3802\r\n   125 |     39.60        68.27   |    2.1622    1.3282\r\n   150 |     47.53        68.67   |    2.1169    1.2775\r\n   175 |     46.93        66.07   |    2.0992    1.3526\r\n   200 |     52.60        68.53   |    2.0492    1.2657\r\n   225 |     55.27        70.67   |    2.0309    1.2157\r\n   250 |     62.13        71.80   |    1.9678    1.1786\r\n   275 |     63.13        72.73   |    1.9374    1.1520\r\n   300 |     65.47        72.40   |    1.8779    1.2167\r\n   325 |     66.00        70.53   |    1.8833    1.2285\r\n   350 |     67.80        69.67   |    1.8193    1.2463\r\n<\/pre>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing text_cell rendered\">\n<div class=\"inner_cell\">\n<div class=\"text_cell_render border-box-sizing rendered_html\">\n<p>\u7d50\u679c\u3092\u30b0\u30e9\u30d5\u3067\u63cf\u753b\u3057\u307e\u3059\u3002<\/p>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing code_cell rendered\">\n<div class=\"input\">\n<div class=\"inner_cell\">\n<div class=\"input_area\">\n<div class=\"highlight hl-python\">\n<pre><span><\/span><span class=\"n\">plot_results_vs_train_size<\/span><span class=\"p\">(<\/span><span class=\"n\">train_sizes<\/span><span class=\"p\">,<\/span>\r\n                           <span class=\"n\">acc_mean_fcnn<\/span><span class=\"p\">,<\/span> <span class=\"n\">acc_std_fcnn<\/span><span class=\"p\">,<\/span> <span class=\"n\">ce_mean_fcnn<\/span><span class=\"p\">,<\/span> <span class=\"n\">ce_std_fcnn<\/span><span class=\"p\">,<\/span>\r\n                           <span class=\"n\">acc_mean_nbmf<\/span><span class=\"p\">,<\/span> <span class=\"n\">acc_std_nbmf<\/span><span class=\"p\">,<\/span> <span class=\"n\">ce_mean_nbmf<\/span><span class=\"p\">,<\/span> <span class=\"n\">ce_std_nbmf<\/span><span class=\"p\">)<\/span>\r\n<\/pre>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"output_wrapper\">\n<div class=\"output\">\n<div class=\"output_area\">\n<div class=\"prompt\"><\/div>\n<div class=\"output_png output_subarea\"><img decoding=\"async\" alt=\"No description has been provided for this image\" src=\"data:image\/png;base64,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\" \/><\/div>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing text_cell rendered\">\n<div class=\"inner_cell\">\n<div class=\"text_cell_render border-box-sizing rendered_html\">\n<h3><span class=\"ez-toc-section\" id=\"%E7%89%B9%E5%BE%B4%E6%95%B0%E3%81%AB%E5%AF%BE%E3%81%99%E3%82%8B%E7%B2%BE%E5%BA%A6%E3%81%A8%E4%BA%A4%E5%B7%AE%E3%82%A8%E3%83%B3%E3%83%88%E3%83%AD%E3%83%94%E3%83%BC%E8%AA%A4%E5%B7%AE%E3%81%A7NBMF%E3%81%A8FCNN%E3%81%AE%E6%AF%94%E8%BC%83%E3%82%92%E8%A1%8C%E3%81%86%E5%AE%9F%E9%A8%93\"><\/span>\u7279\u5fb4\u6570\u306b\u5bfe\u3059\u308b\u7cbe\u5ea6\u3068\u4ea4\u5dee\u30a8\u30f3\u30c8\u30ed\u30d4\u30fc\u8aa4\u5dee\u3067NBMF\u3068FCNN\u306e\u6bd4\u8f03\u3092\u884c\u3046\u5b9f\u9a13<span class=\"ez-toc-section-end\"><\/span><\/h3>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing text_cell rendered\">\n<div class=\"inner_cell\">\n<div class=\"text_cell_render border-box-sizing rendered_html\">\n<p>\u6b21\u306b\u3001\u7279\u5fb4\u6570\u309210\u304b\u3089100\u307e\u306710\u305a\u3064\u5909\u5316\u3055\u305b\u305f\u5834\u5408\u306e\u7cbe\u5ea6\u3068\u4ea4\u5dee\u30a8\u30f3\u30c8\u30ed\u30d4\u30fc\u8aa4\u5dee\u306b\u3064\u3044\u3066\u3001NBMF\u3068FCNN\u3067\u6bd4\u8f03\u3059\u308b\u5b9f\u9a13\u3092\u884c\u3044\u307e\u3059\u3002\u307e\u305a\u306f\u3001\u30d1\u30e9\u30e1\u30fc\u30bf\u306e\u8a2d\u5b9a\u3067\u3059\u3002<\/p>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing code_cell rendered\">\n<div class=\"input\">\n<div class=\"inner_cell\">\n<div class=\"input_area\">\n<div class=\"highlight hl-python\">\n<pre><span><\/span><span class=\"c1\"># \u5171\u901a\u30d1\u30e9\u30e1\u30fc\u30bf<\/span>\r\n<span class=\"n\">epochs<\/span> <span class=\"o\">=<\/span> <span class=\"mi\">10<\/span>\r\n<span class=\"n\">num_repeats<\/span> <span class=\"o\">=<\/span> <span class=\"mi\">3<\/span>\r\n<span class=\"n\">m_train<\/span> <span class=\"o\">=<\/span> <span class=\"mi\">150<\/span>\r\n<span class=\"n\">m_test<\/span> <span class=\"o\">=<\/span> <span class=\"mi\">500<\/span>\r\n<span class=\"n\">seed<\/span> <span class=\"o\">=<\/span> <span class=\"mi\">0<\/span>\r\n<span class=\"n\">k_list<\/span> <span class=\"o\">=<\/span> <span class=\"nb\">list<\/span><span class=\"p\">(<\/span><span class=\"nb\">range<\/span><span class=\"p\">(<\/span><span class=\"mi\">10<\/span><span class=\"p\">,<\/span> <span class=\"mi\">101<\/span><span class=\"p\">,<\/span> <span class=\"mi\">10<\/span><span class=\"p\">))<\/span>\r\n\r\n<span class=\"c1\"># --- FCNN\u30d1\u30e9\u30e1\u30fc\u30bf ---<\/span>\r\n<span class=\"n\">lr<\/span> <span class=\"o\">=<\/span> <span class=\"mf\">0.0002<\/span>\r\n<span class=\"n\">batch_size<\/span> <span class=\"o\">=<\/span> <span class=\"mi\">32<\/span>\r\n\r\n<span class=\"c1\"># --- NBMF\u30d1\u30e9\u30e1\u30fc\u30bf ---<\/span>\r\n<span class=\"n\">g<\/span> <span class=\"o\">=<\/span> <span class=\"mf\">9.0<\/span>\r\n<span class=\"n\">alpha<\/span> <span class=\"o\">=<\/span> <span class=\"mf\">1e-4<\/span>\r\n<span class=\"n\">lr_W<\/span> <span class=\"o\">=<\/span> <span class=\"mf\">0.01<\/span>\r\n<span class=\"n\">l1<\/span> <span class=\"o\">=<\/span> <span class=\"mi\">0<\/span>\r\n<span class=\"c1\"># sampler = oj.SASampler()<\/span>\r\n<span class=\"n\">sampler<\/span> <span class=\"o\">=<\/span> <span class=\"n\">neal<\/span><span class=\"o\">.<\/span><span class=\"n\">SimulatedAnnealingSampler<\/span><span class=\"p\">()<\/span>\r\n<\/pre>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing text_cell rendered\">\n<div class=\"inner_cell\">\n<div class=\"text_cell_render border-box-sizing rendered_html\">\n<p>\u6b21\u306b\u5b66\u7fd2\u30c7\u30fc\u30bf\u3068\u30c6\u30b9\u30c8\u30c7\u30fc\u30bf\u306e\u8aad\u307f\u8fbc\u307f\u3092\u884c\u3044\u307e\u3059\u3002<\/p>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing code_cell rendered\">\n<div class=\"input\">\n<div class=\"inner_cell\">\n<div class=\"input_area\">\n<div class=\"highlight hl-python\">\n<pre><span><\/span><span class=\"n\">X_train<\/span><span class=\"p\">,<\/span> <span class=\"n\">y_train<\/span><span class=\"p\">,<\/span> <span class=\"n\">X_test<\/span><span class=\"p\">,<\/span> <span class=\"n\">y_test<\/span> <span class=\"o\">=<\/span> <span class=\"n\">load_mnist_data<\/span><span class=\"p\">(<\/span><span class=\"n\">m_train<\/span><span class=\"p\">,<\/span> <span class=\"n\">m_test<\/span><span class=\"p\">,<\/span> <span class=\"n\">seed<\/span><span class=\"o\">=<\/span><span class=\"n\">seed<\/span><span class=\"p\">)<\/span>\r\n<\/pre>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing text_cell rendered\">\n<div class=\"inner_cell\">\n<div class=\"text_cell_render border-box-sizing rendered_html\">\n<p>\u305d\u308c\u3067\u306f\u5b9f\u9a13\u3092\u884c\u3044\u307e\u3059\u3002<\/p>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing code_cell rendered\">\n<div class=\"input\">\n<div class=\"inner_cell\">\n<div class=\"input_area\">\n<div class=\"highlight hl-python\">\n<pre><span><\/span><span class=\"n\">acc_mean_fcnn_list<\/span><span class=\"p\">,<\/span> <span class=\"n\">acc_std_fcnn_list<\/span> <span class=\"o\">=<\/span> <span class=\"p\">[],<\/span> <span class=\"p\">[]<\/span>\r\n<span class=\"n\">ce_mean_fcnn_list<\/span><span class=\"p\">,<\/span> <span class=\"n\">ce_std_fcnn_list<\/span> <span class=\"o\">=<\/span> <span class=\"p\">[],<\/span> <span class=\"p\">[]<\/span>\r\n<span class=\"n\">acc_mean_nbmf_list<\/span><span class=\"p\">,<\/span> <span class=\"n\">acc_std_nbmf_list<\/span> <span class=\"o\">=<\/span> <span class=\"p\">[],<\/span> <span class=\"p\">[]<\/span>\r\n<span class=\"n\">ce_mean_nbmf_list<\/span><span class=\"p\">,<\/span> <span class=\"n\">ce_std_nbmf_list<\/span> <span class=\"o\">=<\/span> <span class=\"p\">[],<\/span> <span class=\"p\">[]<\/span>\r\n\r\n<span class=\"k\">for<\/span> <span class=\"n\">k<\/span> <span class=\"ow\">in<\/span> <span class=\"n\">k_list<\/span><span class=\"p\">:<\/span>\r\n    <span class=\"nb\">print<\/span><span class=\"p\">(<\/span><span class=\"sa\">f<\/span><span class=\"s2\">\"<\/span><span class=\"se\">\\n<\/span><span class=\"s2\">--- k = <\/span><span class=\"si\">{<\/span><span class=\"n\">k<\/span><span class=\"si\">}<\/span><span class=\"s2\"> ---\"<\/span><span class=\"p\">)<\/span>\r\n\r\n    <span class=\"c1\"># ---- FCNN ----<\/span>\r\n    <span class=\"n\">accs_fcnn<\/span><span class=\"p\">,<\/span> <span class=\"n\">ces_fcnn<\/span> <span class=\"o\">=<\/span> <span class=\"p\">[],<\/span> <span class=\"p\">[]<\/span>\r\n    <span class=\"k\">for<\/span> <span class=\"n\">s<\/span> <span class=\"ow\">in<\/span> <span class=\"nb\">range<\/span><span class=\"p\">(<\/span><span class=\"n\">num_repeats<\/span><span class=\"p\">):<\/span>\r\n        <span class=\"n\">acc_fcnn<\/span><span class=\"p\">,<\/span> <span class=\"n\">ce_fcnn<\/span> <span class=\"o\">=<\/span> <span class=\"n\">train_fcnn_once<\/span><span class=\"p\">(<\/span>\r\n            <span class=\"n\">X_train<\/span><span class=\"p\">,<\/span> <span class=\"n\">y_train<\/span><span class=\"p\">,<\/span> <span class=\"n\">X_test<\/span><span class=\"p\">,<\/span> <span class=\"n\">y_test<\/span><span class=\"p\">,<\/span>\r\n            <span class=\"n\">hidden_dim<\/span><span class=\"o\">=<\/span><span class=\"n\">k<\/span><span class=\"p\">,<\/span> <span class=\"n\">lr<\/span><span class=\"o\">=<\/span><span class=\"n\">lr<\/span><span class=\"p\">,<\/span> <span class=\"n\">epochs<\/span><span class=\"o\">=<\/span><span class=\"n\">epochs<\/span><span class=\"p\">,<\/span>\r\n            <span class=\"n\">batch_size<\/span><span class=\"o\">=<\/span><span class=\"n\">batch_size<\/span><span class=\"p\">,<\/span> <span class=\"n\">seed<\/span><span class=\"o\">=<\/span><span class=\"n\">s<\/span>\r\n        <span class=\"p\">)<\/span>\r\n        <span class=\"n\">accs_fcnn<\/span><span class=\"o\">.<\/span><span class=\"n\">append<\/span><span class=\"p\">(<\/span><span class=\"n\">acc_fcnn<\/span><span class=\"p\">[<\/span><span class=\"o\">-<\/span><span class=\"mi\">1<\/span><span class=\"p\">])<\/span>\r\n        <span class=\"n\">ces_fcnn<\/span><span class=\"o\">.<\/span><span class=\"n\">append<\/span><span class=\"p\">(<\/span><span class=\"n\">ce_fcnn<\/span><span class=\"p\">[<\/span><span class=\"o\">-<\/span><span class=\"mi\">1<\/span><span class=\"p\">])<\/span>\r\n\r\n    <span class=\"n\">acc_mean_fcnn_list<\/span><span class=\"o\">.<\/span><span class=\"n\">append<\/span><span class=\"p\">(<\/span><span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">mean<\/span><span class=\"p\">(<\/span><span class=\"n\">accs_fcnn<\/span><span class=\"p\">))<\/span>\r\n    <span class=\"n\">acc_std_fcnn_list<\/span><span class=\"o\">.<\/span><span class=\"n\">append<\/span><span class=\"p\">(<\/span><span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">std<\/span><span class=\"p\">(<\/span><span class=\"n\">accs_fcnn<\/span><span class=\"p\">))<\/span>\r\n    <span class=\"n\">ce_mean_fcnn_list<\/span><span class=\"o\">.<\/span><span class=\"n\">append<\/span><span class=\"p\">(<\/span><span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">mean<\/span><span class=\"p\">(<\/span><span class=\"n\">ces_fcnn<\/span><span class=\"p\">))<\/span>\r\n    <span class=\"n\">ce_std_fcnn_list<\/span><span class=\"o\">.<\/span><span class=\"n\">append<\/span><span class=\"p\">(<\/span><span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">std<\/span><span class=\"p\">(<\/span><span class=\"n\">ces_fcnn<\/span><span class=\"p\">))<\/span>\r\n\r\n    <span class=\"c1\"># ---- NBMF ----<\/span>\r\n    <span class=\"n\">accs_nbmf<\/span><span class=\"p\">,<\/span> <span class=\"n\">ces_nbmf<\/span> <span class=\"o\">=<\/span> <span class=\"p\">[],<\/span> <span class=\"p\">[]<\/span>\r\n    <span class=\"k\">for<\/span> <span class=\"n\">s<\/span> <span class=\"ow\">in<\/span> <span class=\"nb\">range<\/span><span class=\"p\">(<\/span><span class=\"n\">num_repeats<\/span><span class=\"p\">):<\/span>\r\n        <span class=\"nb\">print<\/span><span class=\"p\">(<\/span><span class=\"sa\">f<\/span><span class=\"s2\">\"  --- NBMF \u5b9f\u9a13 <\/span><span class=\"si\">{<\/span><span class=\"n\">s<\/span><span class=\"o\">+<\/span><span class=\"mi\">1<\/span><span class=\"si\">}<\/span><span class=\"s2\">\/3 ---\"<\/span><span class=\"p\">)<\/span>\r\n        <span class=\"n\">acc_nbmf<\/span><span class=\"p\">,<\/span> <span class=\"n\">ce_nbmf<\/span><span class=\"p\">,<\/span> <span class=\"n\">W_final<\/span><span class=\"p\">,<\/span> <span class=\"n\">H_final<\/span><span class=\"p\">,<\/span> <span class=\"n\">y_pred<\/span><span class=\"p\">,<\/span> <span class=\"n\">H_test<\/span> <span class=\"o\">=<\/span> <span class=\"n\">train_nbmf_once<\/span><span class=\"p\">(<\/span>\r\n            <span class=\"n\">X_train<\/span><span class=\"p\">,<\/span> <span class=\"n\">y_train<\/span><span class=\"p\">,<\/span> <span class=\"n\">X_test<\/span><span class=\"p\">,<\/span> <span class=\"n\">y_test<\/span><span class=\"p\">,<\/span>\r\n            <span class=\"n\">k<\/span><span class=\"o\">=<\/span><span class=\"n\">k<\/span><span class=\"p\">,<\/span> <span class=\"n\">epochs<\/span><span class=\"o\">=<\/span><span class=\"n\">epochs<\/span><span class=\"p\">,<\/span> <span class=\"n\">alpha<\/span><span class=\"o\">=<\/span><span class=\"n\">alpha<\/span><span class=\"p\">,<\/span> <span class=\"n\">lr_W<\/span><span class=\"o\">=<\/span><span class=\"n\">lr_W<\/span><span class=\"p\">,<\/span> <span class=\"n\">g<\/span><span class=\"o\">=<\/span><span class=\"n\">g<\/span><span class=\"p\">,<\/span><span class=\"n\">l1<\/span><span class=\"o\">=<\/span><span class=\"n\">l1<\/span><span class=\"p\">,<\/span>\r\n            <span class=\"n\">seed<\/span><span class=\"o\">=<\/span><span class=\"n\">s<\/span><span class=\"p\">,<\/span> <span class=\"n\">num_reads<\/span><span class=\"o\">=<\/span><span class=\"mi\">50<\/span><span class=\"p\">,<\/span> <span class=\"n\">verbose<\/span><span class=\"o\">=<\/span><span class=\"kc\">False<\/span><span class=\"p\">,<\/span> <span class=\"n\">sampler<\/span><span class=\"o\">=<\/span><span class=\"n\">sampler<\/span>\r\n        <span class=\"p\">)<\/span>\r\n        <span class=\"n\">accs_nbmf<\/span><span class=\"o\">.<\/span><span class=\"n\">append<\/span><span class=\"p\">(<\/span><span class=\"n\">acc_nbmf<\/span><span class=\"p\">[<\/span><span class=\"o\">-<\/span><span class=\"mi\">1<\/span><span class=\"p\">])<\/span>\r\n        <span class=\"n\">ces_nbmf<\/span><span class=\"o\">.<\/span><span class=\"n\">append<\/span><span class=\"p\">(<\/span><span class=\"n\">ce_nbmf<\/span><span class=\"p\">[<\/span><span class=\"o\">-<\/span><span class=\"mi\">1<\/span><span class=\"p\">])<\/span>\r\n\r\n    <span class=\"n\">acc_mean_nbmf_list<\/span><span class=\"o\">.<\/span><span class=\"n\">append<\/span><span class=\"p\">(<\/span><span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">mean<\/span><span class=\"p\">(<\/span><span class=\"n\">accs_nbmf<\/span><span class=\"p\">))<\/span>\r\n    <span class=\"n\">acc_std_nbmf_list<\/span><span class=\"o\">.<\/span><span class=\"n\">append<\/span><span class=\"p\">(<\/span><span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">std<\/span><span class=\"p\">(<\/span><span class=\"n\">accs_nbmf<\/span><span class=\"p\">))<\/span>\r\n    <span class=\"n\">ce_mean_nbmf_list<\/span><span class=\"o\">.<\/span><span class=\"n\">append<\/span><span class=\"p\">(<\/span><span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">mean<\/span><span class=\"p\">(<\/span><span class=\"n\">ces_nbmf<\/span><span class=\"p\">))<\/span>\r\n    <span class=\"n\">ce_std_nbmf_list<\/span><span class=\"o\">.<\/span><span class=\"n\">append<\/span><span class=\"p\">(<\/span><span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">std<\/span><span class=\"p\">(<\/span><span class=\"n\">ces_nbmf<\/span><span class=\"p\">))<\/span>\r\n\r\n<span class=\"c1\"># numpy\u914d\u5217\u5316<\/span>\r\n<span class=\"n\">acc_mean_fcnn<\/span> <span class=\"o\">=<\/span> <span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">array<\/span><span class=\"p\">(<\/span><span class=\"n\">acc_mean_fcnn_list<\/span><span class=\"p\">)<\/span>\r\n<span class=\"n\">acc_std_fcnn<\/span> <span class=\"o\">=<\/span> <span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">array<\/span><span class=\"p\">(<\/span><span class=\"n\">acc_std_fcnn_list<\/span><span class=\"p\">)<\/span>\r\n<span class=\"n\">ce_mean_fcnn<\/span> <span class=\"o\">=<\/span> <span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">array<\/span><span class=\"p\">(<\/span><span class=\"n\">ce_mean_fcnn_list<\/span><span class=\"p\">)<\/span>\r\n<span class=\"n\">ce_std_fcnn<\/span> <span class=\"o\">=<\/span> <span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">array<\/span><span class=\"p\">(<\/span><span class=\"n\">ce_std_fcnn_list<\/span><span class=\"p\">)<\/span>\r\n<span class=\"n\">acc_mean_nbmf<\/span> <span class=\"o\">=<\/span> <span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">array<\/span><span class=\"p\">(<\/span><span class=\"n\">acc_mean_nbmf_list<\/span><span class=\"p\">)<\/span>\r\n<span class=\"n\">acc_std_nbmf<\/span> <span class=\"o\">=<\/span> <span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">array<\/span><span class=\"p\">(<\/span><span class=\"n\">acc_std_nbmf_list<\/span><span class=\"p\">)<\/span>\r\n<span class=\"n\">ce_mean_nbmf<\/span> <span class=\"o\">=<\/span> <span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">array<\/span><span class=\"p\">(<\/span><span class=\"n\">ce_mean_nbmf_list<\/span><span class=\"p\">)<\/span>\r\n<span class=\"n\">ce_std_nbmf<\/span> <span class=\"o\">=<\/span> <span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">array<\/span><span class=\"p\">(<\/span><span class=\"n\">ce_std_nbmf_list<\/span><span class=\"p\">)<\/span>\r\n<\/pre>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<details>\n<summary style=\"cursor: pointer; font-weight: bold; padding: 10px; background-color: #f0f0f0; border-radius: 5px;\">\n\u5b9f\u9a13\u7d50\u679c\u3092\u8868\u793a\u3059\u308b\uff08\u3053\u3053\u3092\u30af\u30ea\u30c3\u30af\uff09<br \/>\n<\/summary>\n<div class=\"output_wrapper\">\n<div class=\"output\">\n<div class=\"output_area\">\n<div class=\"prompt\"><\/div>\n<div class=\"output_subarea output_stream output_stdout output_text\">\n<pre>--- k = 10 ---\r\n  --- NBMF \u5b9f\u9a13 1\/3 ---\r\n[Epoch 01] TestAcc=18.80%  CE=2.2292\r\n[Epoch 02] TestAcc=36.60%  CE=2.1295\r\n[Epoch 03] TestAcc=36.80%  CE=2.1312\r\n[Epoch 04] TestAcc=36.80%  CE=2.1273\r\n[Epoch 05] TestAcc=38.00%  CE=2.1253\r\n[Epoch 06] TestAcc=37.80%  CE=2.1286\r\n[Epoch 07] TestAcc=37.40%  CE=2.1278\r\n[Epoch 08] TestAcc=36.60%  CE=2.1325\r\n[Epoch 09] TestAcc=36.80%  CE=2.1285\r\n[Epoch 10] TestAcc=38.00%  CE=2.1274\r\n  --- NBMF \u5b9f\u9a13 2\/3 ---\r\n[Epoch 01] TestAcc=25.00%  CE=2.2475\r\n[Epoch 02] TestAcc=41.20%  CE=2.1302\r\n[Epoch 03] TestAcc=41.40%  CE=2.1262\r\n[Epoch 04] TestAcc=42.00%  CE=2.1261\r\n[Epoch 05] TestAcc=40.20%  CE=2.1270\r\n[Epoch 06] TestAcc=41.40%  CE=2.1275\r\n[Epoch 07] TestAcc=42.20%  CE=2.1253\r\n[Epoch 08] TestAcc=41.80%  CE=2.1278\r\n[Epoch 09] TestAcc=42.80%  CE=2.1217\r\n[Epoch 10] TestAcc=39.80%  CE=2.1332\r\n  --- NBMF \u5b9f\u9a13 3\/3 ---\r\n[Epoch 01] TestAcc=27.20%  CE=2.2351\r\n[Epoch 02] TestAcc=35.60%  CE=2.1493\r\n[Epoch 03] TestAcc=35.40%  CE=2.1403\r\n[Epoch 04] TestAcc=35.60%  CE=2.1329\r\n[Epoch 05] TestAcc=36.00%  CE=2.1281\r\n[Epoch 06] TestAcc=34.20%  CE=2.1308\r\n[Epoch 07] TestAcc=34.20%  CE=2.1335\r\n[Epoch 08] TestAcc=35.00%  CE=2.1279\r\n[Epoch 09] TestAcc=36.80%  CE=2.1271\r\n[Epoch 10] TestAcc=35.20%  CE=2.1295\r\n\r\n--- k = 20 ---\r\n  --- NBMF \u5b9f\u9a13 1\/3 ---\r\n[Epoch 01] TestAcc=32.20%  CE=2.1848\r\n[Epoch 02] TestAcc=49.80%  CE=1.9619\r\n[Epoch 03] TestAcc=58.20%  CE=1.7983\r\n[Epoch 04] TestAcc=60.00%  CE=1.7420\r\n[Epoch 05] TestAcc=61.20%  CE=1.7315\r\n[Epoch 06] TestAcc=60.80%  CE=1.7424\r\n[Epoch 07] TestAcc=60.00%  CE=1.7357\r\n[Epoch 08] TestAcc=60.60%  CE=1.7334\r\n[Epoch 09] TestAcc=59.80%  CE=1.7481\r\n[Epoch 10] TestAcc=61.00%  CE=1.7356\r\n  --- NBMF \u5b9f\u9a13 2\/3 ---\r\n[Epoch 01] TestAcc=34.60%  CE=2.1562\r\n[Epoch 02] TestAcc=43.60%  CE=1.9698\r\n[Epoch 03] TestAcc=50.20%  CE=1.8126\r\n[Epoch 04] TestAcc=51.80%  CE=1.7910\r\n[Epoch 05] TestAcc=52.40%  CE=1.7645\r\n[Epoch 06] TestAcc=51.60%  CE=1.7670\r\n[Epoch 07] TestAcc=52.00%  CE=1.7599\r\n[Epoch 08] TestAcc=52.40%  CE=1.7665\r\n[Epoch 09] TestAcc=51.40%  CE=1.7608\r\n[Epoch 10] TestAcc=52.60%  CE=1.7637\r\n  --- NBMF \u5b9f\u9a13 3\/3 ---\r\n[Epoch 01] TestAcc=28.80%  CE=2.1678\r\n[Epoch 02] TestAcc=48.80%  CE=1.9542\r\n[Epoch 03] TestAcc=50.60%  CE=1.9174\r\n[Epoch 04] TestAcc=57.00%  CE=1.7710\r\n[Epoch 05] TestAcc=59.00%  CE=1.7698\r\n[Epoch 06] TestAcc=57.60%  CE=1.7766\r\n[Epoch 07] TestAcc=57.40%  CE=1.7741\r\n[Epoch 08] TestAcc=58.00%  CE=1.7734\r\n[Epoch 09] TestAcc=59.20%  CE=1.7711\r\n[Epoch 10] TestAcc=58.80%  CE=1.7762\r\n\r\n--- k = 30 ---\r\n  --- NBMF \u5b9f\u9a13 1\/3 ---\r\n[Epoch 01] TestAcc=17.60%  CE=2.2508\r\n[Epoch 02] TestAcc=63.60%  CE=1.6157\r\n[Epoch 03] TestAcc=66.20%  CE=1.4610\r\n[Epoch 04] TestAcc=65.00%  CE=1.4679\r\n[Epoch 05] TestAcc=66.60%  CE=1.4570\r\n[Epoch 06] TestAcc=66.80%  CE=1.4358\r\n[Epoch 07] TestAcc=65.60%  CE=1.4258\r\n[Epoch 08] TestAcc=67.40%  CE=1.4199\r\n[Epoch 09] TestAcc=66.60%  CE=1.4198\r\n[Epoch 10] TestAcc=66.40%  CE=1.4209\r\n  --- NBMF \u5b9f\u9a13 2\/3 ---\r\n[Epoch 01] TestAcc=40.00%  CE=2.1410\r\n[Epoch 02] TestAcc=61.00%  CE=1.6448\r\n[Epoch 03] TestAcc=66.40%  CE=1.4739\r\n[Epoch 04] TestAcc=67.80%  CE=1.4596\r\n[Epoch 05] TestAcc=67.60%  CE=1.4588\r\n[Epoch 06] TestAcc=67.20%  CE=1.4542\r\n[Epoch 07] TestAcc=67.80%  CE=1.4593\r\n[Epoch 08] TestAcc=67.60%  CE=1.4652\r\n[Epoch 09] TestAcc=66.80%  CE=1.4569\r\n[Epoch 10] TestAcc=67.80%  CE=1.4588\r\n  --- NBMF \u5b9f\u9a13 3\/3 ---\r\n[Epoch 01] TestAcc=32.80%  CE=2.2112\r\n[Epoch 02] TestAcc=62.80%  CE=1.8672\r\n[Epoch 03] TestAcc=63.80%  CE=1.8252\r\n[Epoch 04] TestAcc=64.40%  CE=1.7527\r\n[Epoch 05] TestAcc=64.20%  CE=1.6947\r\n[Epoch 06] TestAcc=65.00%  CE=1.5654\r\n[Epoch 07] TestAcc=66.60%  CE=1.5308\r\n[Epoch 08] TestAcc=67.40%  CE=1.4624\r\n[Epoch 09] TestAcc=68.00%  CE=1.4339\r\n[Epoch 10] TestAcc=68.60%  CE=1.3973\r\n\r\n--- k = 40 ---\r\n  --- NBMF \u5b9f\u9a13 1\/3 ---\r\n[Epoch 01] TestAcc=44.80%  CE=2.0602\r\n[Epoch 02] TestAcc=66.00%  CE=1.5912\r\n[Epoch 03] TestAcc=66.00%  CE=1.4267\r\n[Epoch 04] TestAcc=66.80%  CE=1.3778\r\n[Epoch 05] TestAcc=66.60%  CE=1.3373\r\n[Epoch 06] TestAcc=66.80%  CE=1.3365\r\n[Epoch 07] TestAcc=67.20%  CE=1.3304\r\n[Epoch 08] TestAcc=66.40%  CE=1.3329\r\n[Epoch 09] TestAcc=66.80%  CE=1.3331\r\n[Epoch 10] TestAcc=67.40%  CE=1.3385\r\n  --- NBMF \u5b9f\u9a13 2\/3 ---\r\n[Epoch 01] TestAcc=30.40%  CE=2.2073\r\n[Epoch 02] TestAcc=68.20%  CE=1.4987\r\n[Epoch 03] TestAcc=71.20%  CE=1.3213\r\n[Epoch 04] TestAcc=71.00%  CE=1.3128\r\n[Epoch 05] TestAcc=70.40%  CE=1.3118\r\n[Epoch 06] TestAcc=71.00%  CE=1.3087\r\n[Epoch 07] TestAcc=71.40%  CE=1.2982\r\n[Epoch 08] TestAcc=71.40%  CE=1.3015\r\n[Epoch 09] TestAcc=70.00%  CE=1.2515\r\n[Epoch 10] TestAcc=71.60%  CE=1.2402\r\n  --- NBMF \u5b9f\u9a13 3\/3 ---\r\n[Epoch 01] TestAcc=28.20%  CE=2.2153\r\n[Epoch 02] TestAcc=66.40%  CE=1.4988\r\n[Epoch 03] TestAcc=67.20%  CE=1.2856\r\n[Epoch 04] TestAcc=70.00%  CE=1.2696\r\n[Epoch 05] TestAcc=69.80%  CE=1.2699\r\n[Epoch 06] TestAcc=69.60%  CE=1.2665\r\n[Epoch 07] TestAcc=70.00%  CE=1.2649\r\n[Epoch 08] TestAcc=70.20%  CE=1.2627\r\n[Epoch 09] TestAcc=68.60%  CE=1.2663\r\n[Epoch 10] TestAcc=70.00%  CE=1.2658\r\n\r\n--- k = 50 ---\r\n  --- NBMF \u5b9f\u9a13 1\/3 ---\r\n[Epoch 01] TestAcc=27.40%  CE=2.2379\r\n[Epoch 02] TestAcc=67.00%  CE=1.5335\r\n[Epoch 03] TestAcc=67.60%  CE=1.4103\r\n[Epoch 04] TestAcc=69.40%  CE=1.3626\r\n[Epoch 05] TestAcc=68.40%  CE=1.3619\r\n[Epoch 06] TestAcc=68.60%  CE=1.3643\r\n[Epoch 07] TestAcc=68.40%  CE=1.3696\r\n[Epoch 08] TestAcc=68.80%  CE=1.3674\r\n[Epoch 09] TestAcc=69.00%  CE=1.3642\r\n[Epoch 10] TestAcc=69.20%  CE=1.3591\r\n  --- NBMF \u5b9f\u9a13 2\/3 ---\r\n[Epoch 01] TestAcc=34.80%  CE=2.1904\r\n[Epoch 02] TestAcc=67.20%  CE=1.5414\r\n[Epoch 03] TestAcc=68.40%  CE=1.3333\r\n[Epoch 04] TestAcc=67.40%  CE=1.2690\r\n[Epoch 05] TestAcc=67.80%  CE=1.2234\r\n[Epoch 06] TestAcc=67.40%  CE=1.2184\r\n[Epoch 07] TestAcc=67.00%  CE=1.2246\r\n[Epoch 08] TestAcc=67.60%  CE=1.2253\r\n[Epoch 09] TestAcc=68.20%  CE=1.2206\r\n[Epoch 10] TestAcc=66.80%  CE=1.2255\r\n  --- NBMF \u5b9f\u9a13 3\/3 ---\r\n[Epoch 01] TestAcc=45.20%  CE=2.0497\r\n[Epoch 02] TestAcc=51.80%  CE=1.9646\r\n[Epoch 03] TestAcc=69.80%  CE=1.4615\r\n[Epoch 04] TestAcc=69.00%  CE=1.3100\r\n[Epoch 05] TestAcc=68.00%  CE=1.2748\r\n[Epoch 06] TestAcc=69.00%  CE=1.2483\r\n[Epoch 07] TestAcc=68.80%  CE=1.2431\r\n[Epoch 08] TestAcc=68.80%  CE=1.2402\r\n[Epoch 09] TestAcc=68.40%  CE=1.2415\r\n[Epoch 10] TestAcc=67.40%  CE=1.2591\r\n\r\n--- k = 60 ---\r\n  --- NBMF \u5b9f\u9a13 1\/3 ---\r\n[Epoch 01] TestAcc=19.40%  CE=2.2562\r\n[Epoch 02] TestAcc=67.20%  CE=1.4473\r\n[Epoch 03] TestAcc=69.00%  CE=1.2043\r\n[Epoch 04] TestAcc=66.00%  CE=1.1817\r\n[Epoch 05] TestAcc=67.40%  CE=1.1796\r\n[Epoch 06] TestAcc=66.80%  CE=1.1812\r\n[Epoch 07] TestAcc=66.60%  CE=1.1857\r\n[Epoch 08] TestAcc=67.00%  CE=1.1815\r\n[Epoch 09] TestAcc=66.20%  CE=1.1881\r\n[Epoch 10] TestAcc=67.20%  CE=1.1895\r\n  --- NBMF \u5b9f\u9a13 2\/3 ---\r\n[Epoch 01] TestAcc=33.80%  CE=2.2201\r\n[Epoch 02] TestAcc=60.40%  CE=1.8780\r\n[Epoch 03] TestAcc=63.20%  CE=1.7172\r\n[Epoch 04] TestAcc=68.40%  CE=1.2890\r\n[Epoch 05] TestAcc=70.60%  CE=1.1726\r\n[Epoch 06] TestAcc=68.40%  CE=1.1563\r\n[Epoch 07] TestAcc=69.20%  CE=1.1581\r\n[Epoch 08] TestAcc=69.40%  CE=1.1558\r\n[Epoch 09] TestAcc=69.40%  CE=1.1533\r\n[Epoch 10] TestAcc=68.40%  CE=1.1549\r\n  --- NBMF \u5b9f\u9a13 3\/3 ---\r\n[Epoch 01] TestAcc=26.80%  CE=2.2354\r\n[Epoch 02] TestAcc=67.00%  CE=1.5245\r\n[Epoch 03] TestAcc=67.20%  CE=1.2709\r\n[Epoch 04] TestAcc=67.20%  CE=1.1887\r\n[Epoch 05] TestAcc=67.00%  CE=1.1796\r\n[Epoch 06] TestAcc=67.20%  CE=1.1842\r\n[Epoch 07] TestAcc=67.00%  CE=1.1847\r\n[Epoch 08] TestAcc=66.80%  CE=1.1842\r\n[Epoch 09] TestAcc=66.60%  CE=1.1829\r\n[Epoch 10] TestAcc=67.80%  CE=1.1791\r\n\r\n--- k = 70 ---\r\n  --- NBMF \u5b9f\u9a13 1\/3 ---\r\n[Epoch 01] TestAcc=48.00%  CE=2.0706\r\n[Epoch 02] TestAcc=67.80%  CE=1.5085\r\n[Epoch 03] TestAcc=72.00%  CE=1.2511\r\n[Epoch 04] TestAcc=69.20%  CE=1.1427\r\n[Epoch 05] TestAcc=70.60%  CE=1.0730\r\n[Epoch 06] TestAcc=68.80%  CE=1.0631\r\n[Epoch 07] TestAcc=69.00%  CE=1.0693\r\n[Epoch 08] TestAcc=69.80%  CE=1.0537\r\n[Epoch 09] TestAcc=68.00%  CE=1.0631\r\n[Epoch 10] TestAcc=69.20%  CE=1.0585\r\n  --- NBMF \u5b9f\u9a13 2\/3 ---\r\n[Epoch 01] TestAcc=35.40%  CE=2.1729\r\n[Epoch 02] TestAcc=67.00%  CE=1.4924\r\n[Epoch 03] TestAcc=67.60%  CE=1.2757\r\n[Epoch 04] TestAcc=69.20%  CE=1.2426\r\n[Epoch 05] TestAcc=69.00%  CE=1.1985\r\n[Epoch 06] TestAcc=68.80%  CE=1.1951\r\n[Epoch 07] TestAcc=69.20%  CE=1.1962\r\n[Epoch 08] TestAcc=70.40%  CE=1.1919\r\n[Epoch 09] TestAcc=69.80%  CE=1.1920\r\n[Epoch 10] TestAcc=70.00%  CE=1.1938\r\n  --- NBMF \u5b9f\u9a13 3\/3 ---\r\n[Epoch 01] TestAcc=25.80%  CE=2.2353\r\n[Epoch 02] TestAcc=71.00%  CE=1.4206\r\n[Epoch 03] TestAcc=71.00%  CE=1.1853\r\n[Epoch 04] TestAcc=71.00%  CE=1.1174\r\n[Epoch 05] TestAcc=71.60%  CE=1.1051\r\n[Epoch 06] TestAcc=70.80%  CE=1.1111\r\n[Epoch 07] TestAcc=71.40%  CE=1.1124\r\n[Epoch 08] TestAcc=71.40%  CE=1.1037\r\n[Epoch 09] TestAcc=72.20%  CE=1.1071\r\n[Epoch 10] TestAcc=71.60%  CE=1.1121\r\n\r\n--- k = 80 ---\r\n  --- NBMF \u5b9f\u9a13 1\/3 ---\r\n[Epoch 01] TestAcc=44.80%  CE=2.1712\r\n[Epoch 02] TestAcc=67.60%  CE=1.5215\r\n[Epoch 03] TestAcc=67.20%  CE=1.2772\r\n[Epoch 04] TestAcc=67.80%  CE=1.1720\r\n[Epoch 05] TestAcc=69.20%  CE=1.0921\r\n[Epoch 06] TestAcc=68.80%  CE=1.0775\r\n[Epoch 07] TestAcc=69.40%  CE=1.0731\r\n[Epoch 08] TestAcc=69.00%  CE=1.0752\r\n[Epoch 09] TestAcc=68.20%  CE=1.0736\r\n[Epoch 10] TestAcc=68.00%  CE=1.0756\r\n  --- NBMF \u5b9f\u9a13 2\/3 ---\r\n[Epoch 01] TestAcc=19.60%  CE=2.2758\r\n[Epoch 02] TestAcc=65.60%  CE=1.4602\r\n[Epoch 03] TestAcc=67.80%  CE=1.2090\r\n[Epoch 04] TestAcc=70.40%  CE=1.1517\r\n[Epoch 05] TestAcc=70.60%  CE=1.1475\r\n[Epoch 06] TestAcc=68.80%  CE=1.1054\r\n[Epoch 07] TestAcc=69.60%  CE=1.0698\r\n[Epoch 08] TestAcc=70.20%  CE=1.0783\r\n[Epoch 09] TestAcc=70.00%  CE=1.0693\r\n[Epoch 10] TestAcc=70.00%  CE=1.0719\r\n  --- NBMF \u5b9f\u9a13 3\/3 ---\r\n[Epoch 01] TestAcc=20.60%  CE=2.2498\r\n[Epoch 02] TestAcc=69.00%  CE=1.3839\r\n[Epoch 03] TestAcc=69.00%  CE=1.1848\r\n[Epoch 04] TestAcc=69.20%  CE=1.1780\r\n[Epoch 05] TestAcc=69.40%  CE=1.1795\r\n[Epoch 06] TestAcc=69.60%  CE=1.1811\r\n[Epoch 07] TestAcc=69.20%  CE=1.1843\r\n[Epoch 08] TestAcc=69.00%  CE=1.1797\r\n[Epoch 09] TestAcc=69.60%  CE=1.1819\r\n[Epoch 10] TestAcc=68.40%  CE=1.1832\r\n\r\n--- k = 90 ---\r\n  --- NBMF \u5b9f\u9a13 1\/3 ---\r\n[Epoch 01] TestAcc=43.40%  CE=2.1199\r\n[Epoch 02] TestAcc=68.60%  CE=1.5444\r\n[Epoch 03] TestAcc=70.60%  CE=1.3059\r\n[Epoch 04] TestAcc=69.20%  CE=1.1874\r\n[Epoch 05] TestAcc=69.00%  CE=1.1896\r\n[Epoch 06] TestAcc=69.00%  CE=1.1921\r\n[Epoch 07] TestAcc=69.40%  CE=1.1786\r\n[Epoch 08] TestAcc=69.20%  CE=1.1853\r\n[Epoch 09] TestAcc=69.20%  CE=1.1840\r\n[Epoch 10] TestAcc=69.00%  CE=1.1831\r\n  --- NBMF \u5b9f\u9a13 2\/3 ---\r\n[Epoch 01] TestAcc=49.60%  CE=2.1088\r\n[Epoch 02] TestAcc=66.40%  CE=1.5018\r\n[Epoch 03] TestAcc=69.40%  CE=1.2488\r\n[Epoch 04] TestAcc=71.60%  CE=1.1482\r\n[Epoch 05] TestAcc=72.20%  CE=1.1141\r\n[Epoch 06] TestAcc=71.40%  CE=1.1146\r\n[Epoch 07] TestAcc=72.00%  CE=1.1187\r\n[Epoch 08] TestAcc=71.00%  CE=1.1212\r\n[Epoch 09] TestAcc=71.00%  CE=1.1200\r\n[Epoch 10] TestAcc=71.80%  CE=1.1212\r\n  --- NBMF \u5b9f\u9a13 3\/3 ---\r\n[Epoch 01] TestAcc=43.00%  CE=2.1106\r\n[Epoch 02] TestAcc=66.60%  CE=1.5392\r\n[Epoch 03] TestAcc=69.00%  CE=1.2897\r\n[Epoch 04] TestAcc=69.60%  CE=1.1392\r\n[Epoch 05] TestAcc=71.60%  CE=1.0617\r\n[Epoch 06] TestAcc=70.60%  CE=1.0283\r\n[Epoch 07] TestAcc=68.20%  CE=1.0378\r\n[Epoch 08] TestAcc=68.80%  CE=1.0311\r\n[Epoch 09] TestAcc=68.60%  CE=1.0441\r\n[Epoch 10] TestAcc=69.40%  CE=1.0375\r\n\r\n--- k = 100 ---\r\n  --- NBMF \u5b9f\u9a13 1\/3 ---\r\n[Epoch 01] TestAcc=22.20%  CE=2.2524\r\n[Epoch 02] TestAcc=69.40%  CE=1.3645\r\n[Epoch 03] TestAcc=72.20%  CE=1.1250\r\n[Epoch 04] TestAcc=70.80%  CE=1.0655\r\n[Epoch 05] TestAcc=69.80%  CE=1.0410\r\n[Epoch 06] TestAcc=69.40%  CE=1.0418\r\n[Epoch 07] TestAcc=70.00%  CE=1.0461\r\n[Epoch 08] TestAcc=71.40%  CE=1.0393\r\n[Epoch 09] TestAcc=69.80%  CE=1.0429\r\n[Epoch 10] TestAcc=71.20%  CE=1.0364\r\n  --- NBMF \u5b9f\u9a13 2\/3 ---\r\n[Epoch 01] TestAcc=48.40%  CE=2.1074\r\n[Epoch 02] TestAcc=65.60%  CE=1.5312\r\n[Epoch 03] TestAcc=68.40%  CE=1.2494\r\n[Epoch 04] TestAcc=69.40%  CE=1.1514\r\n[Epoch 05] TestAcc=67.40%  CE=1.1084\r\n[Epoch 06] TestAcc=68.20%  CE=1.0959\r\n[Epoch 07] TestAcc=66.40%  CE=1.1049\r\n[Epoch 08] TestAcc=67.80%  CE=1.0982\r\n[Epoch 09] TestAcc=68.20%  CE=1.0786\r\n[Epoch 10] TestAcc=67.20%  CE=1.0918\r\n  --- NBMF \u5b9f\u9a13 3\/3 ---\r\n[Epoch 01] TestAcc=36.60%  CE=2.2067\r\n[Epoch 02] TestAcc=68.00%  CE=1.4655\r\n[Epoch 03] TestAcc=70.40%  CE=1.2201\r\n[Epoch 04] TestAcc=71.60%  CE=1.1024\r\n[Epoch 05] TestAcc=71.00%  CE=1.0720\r\n[Epoch 06] TestAcc=70.60%  CE=1.0739\r\n[Epoch 07] TestAcc=71.80%  CE=1.0611\r\n[Epoch 08] TestAcc=70.60%  CE=1.0615\r\n[Epoch 09] TestAcc=71.40%  CE=1.0597\r\n[Epoch 10] TestAcc=71.20%  CE=1.0544\r\n<\/pre>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<\/details>\n<div class=\"cell border-box-sizing text_cell rendered\">\n<div class=\"inner_cell\">\n<div class=\"text_cell_render border-box-sizing rendered_html\">\n<p>\u7d50\u679c\u3092\u8868\u793a\u3057\u307e\u3059\u3002<\/p>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing code_cell rendered\">\n<div class=\"input\">\n<div class=\"inner_cell\">\n<div class=\"input_area\">\n<div class=\"highlight hl-python\">\n<pre><span><\/span><span class=\"nb\">print<\/span><span class=\"p\">(<\/span><span class=\"s2\">\"   k   |  FCNN_Acc(%)  NBMF_Acc(%)  |  FCNN_CE   NBMF_CE\"<\/span><span class=\"p\">)<\/span>\r\n<span class=\"k\">for<\/span> <span class=\"n\">k<\/span><span class=\"p\">,<\/span> <span class=\"n\">af<\/span><span class=\"p\">,<\/span> <span class=\"n\">an<\/span><span class=\"p\">,<\/span> <span class=\"n\">cf<\/span><span class=\"p\">,<\/span> <span class=\"n\">cn<\/span> <span class=\"ow\">in<\/span> <span class=\"nb\">zip<\/span><span class=\"p\">(<\/span><span class=\"n\">k_list<\/span><span class=\"p\">,<\/span> <span class=\"n\">acc_mean_fcnn<\/span><span class=\"p\">,<\/span> <span class=\"n\">acc_mean_nbmf<\/span><span class=\"p\">,<\/span> <span class=\"n\">ce_mean_fcnn<\/span><span class=\"p\">,<\/span> <span class=\"n\">ce_mean_nbmf<\/span><span class=\"p\">):<\/span>\r\n    <span class=\"nb\">print<\/span><span class=\"p\">(<\/span><span class=\"sa\">f<\/span><span class=\"s2\">\"<\/span><span class=\"si\">{<\/span><span class=\"n\">k<\/span><span class=\"si\">:<\/span><span class=\"s2\">5d<\/span><span class=\"si\">}<\/span><span class=\"s2\"> |   <\/span><span class=\"si\">{<\/span><span class=\"n\">af<\/span><span class=\"o\">*<\/span><span class=\"mi\">100<\/span><span class=\"si\">:<\/span><span class=\"s2\">7.2f<\/span><span class=\"si\">}<\/span><span class=\"s2\">      <\/span><span class=\"si\">{<\/span><span class=\"n\">an<\/span><span class=\"o\">*<\/span><span class=\"mi\">100<\/span><span class=\"si\">:<\/span><span class=\"s2\">7.2f<\/span><span class=\"si\">}<\/span><span class=\"s2\">   |  <\/span><span class=\"si\">{<\/span><span class=\"n\">cf<\/span><span class=\"si\">:<\/span><span class=\"s2\">8.4f<\/span><span class=\"si\">}<\/span><span class=\"s2\">  <\/span><span class=\"si\">{<\/span><span class=\"n\">cn<\/span><span class=\"si\">:<\/span><span class=\"s2\">8.4f<\/span><span class=\"si\">}<\/span><span class=\"s2\">\"<\/span><span class=\"p\">)<\/span>\r\n<\/pre>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"output_wrapper\">\n<div class=\"output\">\n<div class=\"output_area\">\n<div class=\"prompt\"><\/div>\n<div class=\"output_subarea output_stream output_stdout output_text\">\n<pre>   k   |  FCNN_Acc(%)  NBMF_Acc(%)  |  FCNN_CE   NBMF_CE\r\n   10 |     21.20        37.67   |    2.2296    2.1300\r\n   20 |     31.27        57.47   |    2.1923    1.7585\r\n   30 |     44.67        67.60   |    2.1245    1.4256\r\n   40 |     47.53        69.67   |    2.1169    1.2815\r\n   50 |     48.20        67.80   |    2.0603    1.2813\r\n   60 |     56.33        67.80   |    2.0662    1.1745\r\n   70 |     59.60        70.27   |    2.0106    1.1215\r\n   80 |     53.00        68.80   |    2.0119    1.1102\r\n   90 |     62.60        70.07   |    1.9880    1.1139\r\n  100 |     59.93        69.87   |    1.9878    1.0609\r\n<\/pre>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing text_cell rendered\">\n<div class=\"inner_cell\">\n<div class=\"text_cell_render border-box-sizing rendered_html\">\n<p>\u7d50\u679c\u3092\u30b0\u30e9\u30d5\u3068\u3057\u3066\u63cf\u753b\u3057\u307e\u3059\u3002<\/p>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing code_cell rendered\">\n<div class=\"input\">\n<div class=\"inner_cell\">\n<div class=\"input_area\">\n<div class=\"highlight hl-python\">\n<pre><span><\/span><span class=\"n\">plot_results_vs_k<\/span><span class=\"p\">(<\/span><span class=\"n\">k_list<\/span><span class=\"p\">,<\/span>\r\n                  <span class=\"n\">acc_mean_fcnn<\/span><span class=\"p\">,<\/span> <span class=\"n\">acc_std_fcnn<\/span><span class=\"p\">,<\/span> <span class=\"n\">ce_mean_fcnn<\/span><span class=\"p\">,<\/span> <span class=\"n\">ce_std_fcnn<\/span><span class=\"p\">,<\/span>\r\n                  <span class=\"n\">acc_mean_nbmf<\/span><span class=\"p\">,<\/span> <span class=\"n\">acc_std_nbmf<\/span><span class=\"p\">,<\/span> <span class=\"n\">ce_mean_nbmf<\/span><span class=\"p\">,<\/span> <span class=\"n\">ce_std_nbmf<\/span><span class=\"p\">)<\/span>\r\n<\/pre>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"output_wrapper\">\n<div class=\"output\">\n<div class=\"output_area\">\n<div class=\"prompt\"><\/div>\n<div class=\"output_png output_subarea\"><img decoding=\"async\" alt=\"No description has been provided for this image\" src=\"data:image\/png;base64,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\" \/><\/div>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing text_cell rendered\">\n<div class=\"inner_cell\">\n<div class=\"text_cell_render border-box-sizing rendered_html\">\n<h3><span class=\"ez-toc-section\" id=\"%E3%82%A8%E3%83%9D%E3%83%83%E3%82%AF%E6%95%B0%E3%81%AB%E5%AF%BE%E3%81%99%E3%82%8B%E7%B2%BE%E5%BA%A6%E3%81%A8%E4%BA%A4%E5%B7%AE%E3%82%A8%E3%83%B3%E3%83%88%E3%83%AD%E3%83%94%E3%83%BC%E8%AA%A4%E5%B7%AE%E3%81%A7NBMF%E3%81%A8FCNN%E3%81%AE%E6%AF%94%E8%BC%83%E3%82%92%E8%A1%8C%E3%81%86%E5%AE%9F%E9%A8%93\"><\/span>\u30a8\u30dd\u30c3\u30af\u6570\u306b\u5bfe\u3059\u308b\u7cbe\u5ea6\u3068\u4ea4\u5dee\u30a8\u30f3\u30c8\u30ed\u30d4\u30fc\u8aa4\u5dee\u3067NBMF\u3068FCNN\u306e\u6bd4\u8f03\u3092\u884c\u3046\u5b9f\u9a13<span class=\"ez-toc-section-end\"><\/span><\/h3>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing text_cell rendered\">\n<div class=\"inner_cell\">\n<div class=\"text_cell_render border-box-sizing rendered_html\">\n<p>\u6700\u5f8c\u306b\u3001\u30a8\u30dd\u30c3\u30af\u6570\u3092100\u306b\u8a2d\u5b9a\u3057, 1\u30a8\u30dd\u30c3\u30af\u3054\u3068\u306e\u7cbe\u5ea6\u3068\u4ea4\u5dee\u30a8\u30f3\u30c8\u30ed\u30d4\u30fc\u8aa4\u5dee\u304b\u3089NBMF\u3068FCNN\u3067\u6bd4\u8f03\u3059\u308b\u5b9f\u9a13\u3092\u884c\u3044\u307e\u3059\u3002\u307e\u305a\u306f\u3001\u30d1\u30e9\u30e1\u30fc\u30bf\u306e\u8a2d\u5b9a\u3067\u3059\u3002<\/p>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing code_cell rendered\">\n<div class=\"input\">\n<div class=\"inner_cell\">\n<div class=\"input_area\">\n<div class=\"highlight hl-python\">\n<pre><span><\/span><span class=\"n\">m_train<\/span> <span class=\"o\">=<\/span> <span class=\"mi\">300<\/span>\r\n<span class=\"n\">m_test<\/span> <span class=\"o\">=<\/span> <span class=\"mi\">500<\/span>\r\n<span class=\"n\">num_repeats<\/span><span class=\"o\">=<\/span><span class=\"mi\">3<\/span>\r\n<span class=\"n\">seed<\/span> <span class=\"o\">=<\/span> <span class=\"mi\">0<\/span>\r\n\r\n<span class=\"c1\"># --- FCNN\u30d1\u30e9\u30e1\u30fc\u30bf ---<\/span>\r\n<span class=\"n\">hidden_dim<\/span> <span class=\"o\">=<\/span> <span class=\"mi\">40<\/span>\r\n<span class=\"n\">fcnn_epochs<\/span> <span class=\"o\">=<\/span> <span class=\"mi\">200<\/span>\r\n<span class=\"n\">lr<\/span> <span class=\"o\">=<\/span> <span class=\"mf\">0.0002<\/span>\r\n<span class=\"n\">batch_size<\/span> <span class=\"o\">=<\/span> <span class=\"mi\">32<\/span>\r\n\r\n<span class=\"c1\"># --- NBMF\u30d1\u30e9\u30e1\u30fc\u30bf ---<\/span>\r\n<span class=\"n\">k<\/span> <span class=\"o\">=<\/span> <span class=\"mi\">40<\/span>\r\n<span class=\"n\">nbmf_epochs<\/span> <span class=\"o\">=<\/span> <span class=\"mi\">100<\/span>\r\n<span class=\"n\">num_reads<\/span><span class=\"o\">=<\/span><span class=\"mi\">50<\/span>\r\n<span class=\"n\">g<\/span> <span class=\"o\">=<\/span> <span class=\"mf\">9.0<\/span>\r\n<span class=\"n\">alpha<\/span> <span class=\"o\">=<\/span> <span class=\"mf\">1e-4<\/span>\r\n<span class=\"n\">lr_W<\/span> <span class=\"o\">=<\/span> <span class=\"mf\">0.01<\/span>\r\n<span class=\"n\">l1<\/span> <span class=\"o\">=<\/span> <span class=\"mi\">0<\/span>\r\n<span class=\"n\">seed<\/span> <span class=\"o\">=<\/span> <span class=\"mi\">0<\/span>\r\n<span class=\"c1\"># sampler = neal.SimulatedAnnealingSampler()<\/span>\r\n<span class=\"n\">sampler<\/span> <span class=\"o\">=<\/span> <span class=\"n\">oj<\/span><span class=\"o\">.<\/span><span class=\"n\">SASampler<\/span><span class=\"p\">()<\/span>\r\n<\/pre>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing text_cell rendered\">\n<div class=\"inner_cell\">\n<div class=\"text_cell_render border-box-sizing rendered_html\">\n<p>\u5b66\u7fd2\u30c7\u30fc\u30bf\u3068\u30c6\u30b9\u30c8\u30c7\u30fc\u30bf\u3092\u8aad\u307f\u8fbc\u307f\u307e\u3059\u3002<\/p>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing code_cell rendered\">\n<div class=\"input\">\n<div class=\"inner_cell\">\n<div class=\"input_area\">\n<div class=\"highlight hl-python\">\n<pre><span><\/span><span class=\"n\">X_train<\/span><span class=\"p\">,<\/span> <span class=\"n\">y_train<\/span><span class=\"p\">,<\/span> <span class=\"n\">X_test<\/span><span class=\"p\">,<\/span> <span class=\"n\">y_test<\/span> <span class=\"o\">=<\/span> <span class=\"n\">load_mnist_data<\/span><span class=\"p\">(<\/span><span class=\"n\">m_train<\/span><span class=\"p\">,<\/span> <span class=\"n\">m_test<\/span><span class=\"p\">,<\/span> <span class=\"n\">seed<\/span><span class=\"o\">=<\/span><span class=\"n\">seed<\/span><span class=\"p\">)<\/span>\r\n<\/pre>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing text_cell rendered\">\n<div class=\"inner_cell\">\n<div class=\"text_cell_render border-box-sizing rendered_html\">\n<p>FCNN\u306e\u5b9f\u9a13\u3092\u884c\u3044\u307e\u3059\u3002<\/p>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing code_cell rendered\">\n<div class=\"input\">\n<div class=\"inner_cell\">\n<div class=\"input_area\">\n<div class=\"highlight hl-python\">\n<pre><span><\/span><span class=\"c1\"># --- FCNN ---<\/span>\r\n<span class=\"n\">accs_fcnn<\/span><span class=\"p\">,<\/span> <span class=\"n\">ces_fcnn<\/span> <span class=\"o\">=<\/span> <span class=\"p\">[],<\/span> <span class=\"p\">[]<\/span>\r\n<span class=\"n\">acc_mean_fcnn<\/span><span class=\"p\">,<\/span> <span class=\"n\">acc_std_fcnn<\/span> <span class=\"o\">=<\/span> <span class=\"p\">[],<\/span> <span class=\"p\">[]<\/span>\r\n<span class=\"n\">ce_mean_fcnn<\/span><span class=\"p\">,<\/span> <span class=\"n\">ce_std_fcnn<\/span> <span class=\"o\">=<\/span> <span class=\"p\">[],<\/span> <span class=\"p\">[]<\/span>\r\n<span class=\"k\">for<\/span> <span class=\"n\">s<\/span> <span class=\"ow\">in<\/span> <span class=\"nb\">range<\/span><span class=\"p\">(<\/span><span class=\"n\">num_repeats<\/span><span class=\"p\">):<\/span>\r\n    <span class=\"nb\">print<\/span><span class=\"p\">(<\/span><span class=\"sa\">f<\/span><span class=\"s2\">\"<\/span><span class=\"se\">\\n<\/span><span class=\"s2\">=== FCNN \u5b9f\u9a13 <\/span><span class=\"si\">{<\/span><span class=\"n\">s<\/span><span class=\"o\">+<\/span><span class=\"mi\">1<\/span><span class=\"si\">}<\/span><span class=\"s2\">\/3 ===\"<\/span><span class=\"p\">)<\/span>\r\n    <span class=\"n\">acc_fcnn<\/span><span class=\"p\">,<\/span> <span class=\"n\">ce_fcnn<\/span> <span class=\"o\">=<\/span> <span class=\"n\">train_fcnn_once<\/span><span class=\"p\">(<\/span>\r\n        <span class=\"n\">X_train<\/span><span class=\"p\">,<\/span> <span class=\"n\">y_train<\/span><span class=\"p\">,<\/span> <span class=\"n\">X_test<\/span><span class=\"p\">,<\/span> <span class=\"n\">y_test<\/span><span class=\"p\">,<\/span>\r\n        <span class=\"n\">hidden_dim<\/span><span class=\"o\">=<\/span><span class=\"n\">hidden_dim<\/span><span class=\"p\">,<\/span> <span class=\"n\">lr<\/span><span class=\"o\">=<\/span><span class=\"n\">lr<\/span><span class=\"p\">,<\/span> <span class=\"n\">epochs<\/span><span class=\"o\">=<\/span><span class=\"n\">fcnn_epochs<\/span><span class=\"p\">,<\/span>\r\n        <span class=\"n\">batch_size<\/span><span class=\"o\">=<\/span><span class=\"n\">batch_size<\/span><span class=\"p\">,<\/span> <span class=\"n\">seed<\/span><span class=\"o\">=<\/span><span class=\"n\">s<\/span>\r\n    <span class=\"p\">)<\/span>\r\n    <span class=\"n\">accs_fcnn<\/span><span class=\"o\">.<\/span><span class=\"n\">append<\/span><span class=\"p\">(<\/span><span class=\"n\">acc_fcnn<\/span><span class=\"p\">)<\/span>\r\n    <span class=\"n\">ces_fcnn<\/span><span class=\"o\">.<\/span><span class=\"n\">append<\/span><span class=\"p\">(<\/span><span class=\"n\">ce_fcnn<\/span><span class=\"p\">)<\/span>\r\n<span class=\"n\">accs_fcnn<\/span> <span class=\"o\">=<\/span> <span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">array<\/span><span class=\"p\">(<\/span><span class=\"n\">accs_fcnn<\/span><span class=\"p\">)<\/span>\r\n<span class=\"n\">ces_fcnn<\/span> <span class=\"o\">=<\/span> <span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">array<\/span><span class=\"p\">(<\/span><span class=\"n\">ces_fcnn<\/span><span class=\"p\">)<\/span>\r\n<span class=\"n\">acc_mean_fcnn<\/span><span class=\"p\">,<\/span> <span class=\"n\">acc_std_fcnn<\/span> <span class=\"o\">=<\/span> <span class=\"n\">accs_fcnn<\/span><span class=\"o\">.<\/span><span class=\"n\">mean<\/span><span class=\"p\">(<\/span><span class=\"n\">axis<\/span><span class=\"o\">=<\/span><span class=\"mi\">0<\/span><span class=\"p\">),<\/span> <span class=\"n\">accs_fcnn<\/span><span class=\"o\">.<\/span><span class=\"n\">std<\/span><span class=\"p\">(<\/span><span class=\"n\">axis<\/span><span class=\"o\">=<\/span><span class=\"mi\">0<\/span><span class=\"p\">)<\/span>\r\n<span class=\"n\">ce_mean_fcnn<\/span><span class=\"p\">,<\/span> <span class=\"n\">ce_std_fcnn<\/span> <span class=\"o\">=<\/span> <span class=\"n\">ces_fcnn<\/span><span class=\"o\">.<\/span><span class=\"n\">mean<\/span><span class=\"p\">(<\/span><span class=\"n\">axis<\/span><span class=\"o\">=<\/span><span class=\"mi\">0<\/span><span class=\"p\">),<\/span> <span class=\"n\">ces_fcnn<\/span><span class=\"o\">.<\/span><span class=\"n\">std<\/span><span class=\"p\">(<\/span><span class=\"n\">axis<\/span><span class=\"o\">=<\/span><span class=\"mi\">0<\/span><span class=\"p\">)<\/span>\r\n<\/pre>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"output_wrapper\">\n<div class=\"output\">\n<div class=\"output_area\">\n<div class=\"prompt\"><\/div>\n<div class=\"output_subarea output_stream output_stdout output_text\">\n<pre>=== FCNN \u5b9f\u9a13 1\/3 ===\r\n\r\n=== FCNN \u5b9f\u9a13 2\/3 ===\r\n\r\n=== FCNN \u5b9f\u9a13 3\/3 ===\r\n<\/pre>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing text_cell rendered\">\n<div class=\"inner_cell\">\n<div class=\"text_cell_render border-box-sizing rendered_html\">\n<p>NBMF\u306e\u5b9f\u9a13\u3092\u884c\u3044\u307e\u3059\u3002<\/p>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing code_cell rendered\">\n<div class=\"input\">\n<div class=\"inner_cell\">\n<div class=\"input_area\">\n<div class=\"highlight hl-python\">\n<pre><span><\/span><span class=\"c1\"># --- NBMF ---<\/span>\r\n<span class=\"n\">accs_nbmf<\/span><span class=\"p\">,<\/span> <span class=\"n\">ces_nbmf<\/span> <span class=\"o\">=<\/span> <span class=\"p\">[],<\/span> <span class=\"p\">[]<\/span>\r\n<span class=\"n\">acc_mean_nbmf<\/span><span class=\"p\">,<\/span> <span class=\"n\">acc_std_nbmf<\/span> <span class=\"o\">=<\/span> <span class=\"p\">[],<\/span> <span class=\"p\">[]<\/span>\r\n<span class=\"n\">ce_mean_nbmf<\/span><span class=\"p\">,<\/span> <span class=\"n\">ce_std_nbmf<\/span> <span class=\"o\">=<\/span> <span class=\"p\">[],<\/span> <span class=\"p\">[]<\/span>\r\n<span class=\"k\">for<\/span> <span class=\"n\">s<\/span> <span class=\"ow\">in<\/span> <span class=\"nb\">range<\/span><span class=\"p\">(<\/span><span class=\"n\">num_repeats<\/span><span class=\"p\">):<\/span>\r\n    <span class=\"nb\">print<\/span><span class=\"p\">(<\/span><span class=\"sa\">f<\/span><span class=\"s2\">\"<\/span><span class=\"se\">\\n<\/span><span class=\"s2\">=== NBMF \u5b9f\u9a13 <\/span><span class=\"si\">{<\/span><span class=\"n\">s<\/span><span class=\"o\">+<\/span><span class=\"mi\">1<\/span><span class=\"si\">}<\/span><span class=\"s2\">\/3 ===\"<\/span><span class=\"p\">)<\/span>\r\n    <span class=\"n\">acc_nbmf<\/span><span class=\"p\">,<\/span> <span class=\"n\">ce_nbmf<\/span><span class=\"p\">,<\/span> <span class=\"n\">W_final<\/span><span class=\"p\">,<\/span> <span class=\"n\">H_final<\/span><span class=\"p\">,<\/span> <span class=\"n\">y_pred<\/span><span class=\"p\">,<\/span> <span class=\"n\">H_test<\/span> <span class=\"o\">=<\/span> <span class=\"n\">train_nbmf_once<\/span><span class=\"p\">(<\/span>\r\n        <span class=\"n\">X_train<\/span><span class=\"p\">,<\/span> <span class=\"n\">y_train<\/span><span class=\"p\">,<\/span> <span class=\"n\">X_test<\/span><span class=\"p\">,<\/span> <span class=\"n\">y_test<\/span><span class=\"p\">,<\/span>\r\n        <span class=\"n\">k<\/span><span class=\"o\">=<\/span><span class=\"n\">k<\/span><span class=\"p\">,<\/span> <span class=\"n\">epochs<\/span><span class=\"o\">=<\/span><span class=\"n\">nbmf_epochs<\/span><span class=\"p\">,<\/span> <span class=\"n\">alpha<\/span><span class=\"o\">=<\/span><span class=\"n\">alpha<\/span><span class=\"p\">,<\/span> <span class=\"n\">lr_W<\/span><span class=\"o\">=<\/span><span class=\"n\">lr_W<\/span><span class=\"p\">,<\/span> <span class=\"n\">g<\/span><span class=\"o\">=<\/span><span class=\"n\">g<\/span><span class=\"p\">,<\/span> <span class=\"n\">l1<\/span><span class=\"o\">=<\/span><span class=\"n\">l1<\/span><span class=\"p\">,<\/span>\r\n        <span class=\"n\">seed<\/span><span class=\"o\">=<\/span><span class=\"n\">s<\/span><span class=\"p\">,<\/span> <span class=\"n\">num_reads<\/span><span class=\"o\">=<\/span><span class=\"n\">num_reads<\/span><span class=\"p\">,<\/span> <span class=\"n\">verbose<\/span><span class=\"o\">=<\/span><span class=\"kc\">True<\/span><span class=\"p\">,<\/span> <span class=\"n\">sampler<\/span><span class=\"o\">=<\/span><span class=\"n\">sampler<\/span>\r\n    <span class=\"p\">)<\/span>\r\n    <span class=\"n\">L<\/span> <span class=\"o\">=<\/span> <span class=\"nb\">len<\/span><span class=\"p\">(<\/span><span class=\"n\">acc_nbmf<\/span><span class=\"p\">)<\/span>\r\n    <span class=\"k\">if<\/span> <span class=\"n\">L<\/span> <span class=\"o\">&lt;<\/span> <span class=\"n\">nbmf_epochs<\/span><span class=\"p\">:<\/span>\r\n        <span class=\"n\">pad_len<\/span> <span class=\"o\">=<\/span> <span class=\"n\">nbmf_epochs<\/span> <span class=\"o\">-<\/span> <span class=\"n\">L<\/span>\r\n        <span class=\"n\">acc_nbmf<\/span> <span class=\"o\">=<\/span> <span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">concatenate<\/span><span class=\"p\">([<\/span><span class=\"n\">acc_nbmf<\/span><span class=\"p\">,<\/span> <span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">full<\/span><span class=\"p\">(<\/span><span class=\"n\">pad_len<\/span><span class=\"p\">,<\/span> <span class=\"n\">acc_nbmf<\/span><span class=\"p\">[<\/span><span class=\"o\">-<\/span><span class=\"mi\">1<\/span><span class=\"p\">])])<\/span>\r\n        <span class=\"n\">ce_nbmf<\/span> <span class=\"o\">=<\/span> <span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">concatenate<\/span><span class=\"p\">([<\/span><span class=\"n\">ce_nbmf<\/span><span class=\"p\">,<\/span> <span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">full<\/span><span class=\"p\">(<\/span><span class=\"n\">pad_len<\/span><span class=\"p\">,<\/span> <span class=\"n\">ce_nbmf<\/span><span class=\"p\">[<\/span><span class=\"o\">-<\/span><span class=\"mi\">1<\/span><span class=\"p\">])])<\/span>\r\n    <span class=\"n\">accs_nbmf<\/span><span class=\"o\">.<\/span><span class=\"n\">append<\/span><span class=\"p\">(<\/span><span class=\"n\">acc_nbmf<\/span><span class=\"p\">[:<\/span><span class=\"n\">nbmf_epochs<\/span><span class=\"p\">])<\/span>\r\n    <span class=\"n\">ces_nbmf<\/span><span class=\"o\">.<\/span><span class=\"n\">append<\/span><span class=\"p\">(<\/span><span class=\"n\">ce_nbmf<\/span><span class=\"p\">[:<\/span><span class=\"n\">nbmf_epochs<\/span><span class=\"p\">])<\/span>\r\n\r\n<span class=\"n\">accs_nbmf<\/span> <span class=\"o\">=<\/span> <span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">array<\/span><span class=\"p\">(<\/span><span class=\"n\">accs_nbmf<\/span><span class=\"p\">)<\/span>\r\n<span class=\"n\">ces_nbmf<\/span> <span class=\"o\">=<\/span> <span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">array<\/span><span class=\"p\">(<\/span><span class=\"n\">ces_nbmf<\/span><span class=\"p\">)<\/span>\r\n<span class=\"n\">acc_mean_nbmf<\/span><span class=\"p\">,<\/span> <span class=\"n\">acc_std_nbmf<\/span> <span class=\"o\">=<\/span> <span class=\"n\">accs_nbmf<\/span><span class=\"o\">.<\/span><span class=\"n\">mean<\/span><span class=\"p\">(<\/span><span class=\"n\">axis<\/span><span class=\"o\">=<\/span><span class=\"mi\">0<\/span><span class=\"p\">),<\/span> <span class=\"n\">accs_nbmf<\/span><span class=\"o\">.<\/span><span class=\"n\">std<\/span><span class=\"p\">(<\/span><span class=\"n\">axis<\/span><span class=\"o\">=<\/span><span class=\"mi\">0<\/span><span class=\"p\">)<\/span>\r\n<span class=\"n\">ce_mean_nbmf<\/span><span class=\"p\">,<\/span> <span class=\"n\">ce_std_nbmf<\/span> <span class=\"o\">=<\/span> <span class=\"n\">ces_nbmf<\/span><span class=\"o\">.<\/span><span class=\"n\">mean<\/span><span class=\"p\">(<\/span><span class=\"n\">axis<\/span><span class=\"o\">=<\/span><span class=\"mi\">0<\/span><span class=\"p\">),<\/span> <span class=\"n\">ces_nbmf<\/span><span class=\"o\">.<\/span><span class=\"n\">std<\/span><span class=\"p\">(<\/span><span class=\"n\">axis<\/span><span class=\"o\">=<\/span><span class=\"mi\">0<\/span><span class=\"p\">)<\/span>\r\n<\/pre>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<details>\n<summary style=\"cursor: pointer; font-weight: bold; padding: 10px; background-color: #f0f0f0; border-radius: 5px;\">\n\u5b9f\u9a13\u7d50\u679c\u3092\u8868\u793a\u3059\u308b\uff08\u3053\u3053\u3092\u30af\u30ea\u30c3\u30af\uff09<br \/>\n<\/summary>\n<div class=\"output_wrapper\">\n<div class=\"output\">\n<div class=\"output_area\">\n<div class=\"prompt\"><\/div>\n<div class=\"output_subarea output_stream output_stdout output_text\">\n<pre>=== NBMF \u5b9f\u9a13 1\/3 ===\r\n[Epoch 01] TestAcc=17.00%  CE=2.2679\r\n[Epoch 02] TestAcc=62.80%  CE=1.6153\r\n[Epoch 03] TestAcc=64.60%  CE=1.3973\r\n[Epoch 04] TestAcc=66.40%  CE=1.2598\r\n[Epoch 05] TestAcc=66.20%  CE=1.2397\r\n[Epoch 06] TestAcc=66.60%  CE=1.2414\r\n[Epoch 07] TestAcc=70.00%  CE=1.1960\r\n[Epoch 08] TestAcc=69.20%  CE=1.2080\r\n[Epoch 09] TestAcc=68.40%  CE=1.2084\r\n[Epoch 10] TestAcc=69.00%  CE=1.2254\r\n[Epoch 11] TestAcc=69.80%  CE=1.2148\r\n[Epoch 12] TestAcc=71.40%  CE=1.2079\r\n[Epoch 13] TestAcc=68.40%  CE=1.1918\r\n[Epoch 14] TestAcc=69.20%  CE=1.2030\r\n[Epoch 15] TestAcc=68.80%  CE=1.2162\r\n[Epoch 16] TestAcc=67.40%  CE=1.2305\r\n[Epoch 17] TestAcc=69.60%  CE=1.2097\r\n[Epoch 18] TestAcc=68.40%  CE=1.2100\r\n[Epoch 19] TestAcc=68.40%  CE=1.1948\r\n[Epoch 20] TestAcc=71.60%  CE=1.1994\r\n[Epoch 21] TestAcc=70.00%  CE=1.2046\r\n[Epoch 22] TestAcc=69.80%  CE=1.2070\r\n[Epoch 23] TestAcc=68.80%  CE=1.2195\r\n[Epoch 24] TestAcc=69.00%  CE=1.2194\r\n[Epoch 25] TestAcc=71.00%  CE=1.2088\r\n[Epoch 26] TestAcc=67.20%  CE=1.2277\r\n[Epoch 27] TestAcc=68.80%  CE=1.2182\r\n[Epoch 28] TestAcc=67.60%  CE=1.1952\r\n[Epoch 29] TestAcc=68.20%  CE=1.2144\r\n[Epoch 30] TestAcc=70.40%  CE=1.1850\r\n[Epoch 31] TestAcc=66.20%  CE=1.2281\r\n[Epoch 32] TestAcc=68.80%  CE=1.2136\r\n[Epoch 33] TestAcc=70.80%  CE=1.1961\r\n[Epoch 34] TestAcc=69.60%  CE=1.2175\r\n[Epoch 35] TestAcc=67.80%  CE=1.1840\r\n[Epoch 36] TestAcc=69.80%  CE=1.2068\r\n[Epoch 37] TestAcc=68.20%  CE=1.2037\r\n[Epoch 38] TestAcc=70.00%  CE=1.1997\r\n[Epoch 39] TestAcc=68.00%  CE=1.2253\r\n[Epoch 40] TestAcc=70.60%  CE=1.1851\r\n[Epoch 41] TestAcc=69.40%  CE=1.2078\r\n[Epoch 42] TestAcc=69.00%  CE=1.2178\r\n[Epoch 43] TestAcc=70.40%  CE=1.2088\r\n[Epoch 44] TestAcc=69.00%  CE=1.1895\r\n[Epoch 45] TestAcc=70.60%  CE=1.1912\r\n[Epoch 46] TestAcc=68.60%  CE=1.2158\r\n[Epoch 47] TestAcc=68.00%  CE=1.2136\r\n[Epoch 48] TestAcc=68.60%  CE=1.2043\r\n[Epoch 49] TestAcc=68.60%  CE=1.1970\r\n[Epoch 50] TestAcc=67.40%  CE=1.2061\r\n[Epoch 51] TestAcc=71.60%  CE=1.1749\r\n[Epoch 52] TestAcc=69.00%  CE=1.1947\r\n[Epoch 53] TestAcc=68.60%  CE=1.1863\r\n[Epoch 54] TestAcc=70.40%  CE=1.1777\r\n[Epoch 55] TestAcc=70.80%  CE=1.1868\r\n[Epoch 56] TestAcc=69.20%  CE=1.2053\r\n[Epoch 57] TestAcc=69.40%  CE=1.2033\r\n[Epoch 58] TestAcc=70.40%  CE=1.1850\r\n[Epoch 59] TestAcc=68.80%  CE=1.1924\r\n[Epoch 60] TestAcc=70.40%  CE=1.1822\r\n[Epoch 61] TestAcc=70.80%  CE=1.1986\r\n[Epoch 62] TestAcc=69.60%  CE=1.1996\r\n[Epoch 63] TestAcc=69.20%  CE=1.2171\r\n[Epoch 64] TestAcc=70.20%  CE=1.1958\r\n[Epoch 65] TestAcc=69.60%  CE=1.1848\r\n[Epoch 66] TestAcc=71.80%  CE=1.1709\r\n[Epoch 67] TestAcc=70.60%  CE=1.1864\r\n[Epoch 68] TestAcc=68.00%  CE=1.2165\r\n[Epoch 69] TestAcc=68.40%  CE=1.1877\r\n[Epoch 70] TestAcc=71.20%  CE=1.1737\r\n[Epoch 71] TestAcc=71.40%  CE=1.1737\r\n[Epoch 72] TestAcc=70.40%  CE=1.2061\r\n[Epoch 73] TestAcc=69.00%  CE=1.2029\r\n[Epoch 74] TestAcc=70.00%  CE=1.1741\r\n[Epoch 75] TestAcc=69.60%  CE=1.1947\r\n[Epoch 76] TestAcc=66.20%  CE=1.1984\r\n[Epoch 77] TestAcc=69.00%  CE=1.2000\r\n[Epoch 78] TestAcc=69.40%  CE=1.1800\r\n[Epoch 79] TestAcc=69.20%  CE=1.1972\r\n[Epoch 80] TestAcc=69.80%  CE=1.1820\r\n[Epoch 81] TestAcc=69.00%  CE=1.1807\r\n[Epoch 82] TestAcc=68.60%  CE=1.2031\r\n[Epoch 83] TestAcc=72.00%  CE=1.1574\r\n[Epoch 84] TestAcc=69.60%  CE=1.1743\r\n[Epoch 85] TestAcc=70.00%  CE=1.1918\r\n[Epoch 86] TestAcc=70.00%  CE=1.1600\r\n[Epoch 87] TestAcc=72.40%  CE=1.1289\r\n[Epoch 88] TestAcc=71.60%  CE=1.1280\r\n[Epoch 89] TestAcc=70.20%  CE=1.1711\r\n[Epoch 90] TestAcc=71.80%  CE=1.1454\r\n[Epoch 91] TestAcc=70.20%  CE=1.1601\r\n[Epoch 92] TestAcc=70.20%  CE=1.1548\r\n[Epoch 93] TestAcc=71.40%  CE=1.1648\r\n[Epoch 94] TestAcc=71.60%  CE=1.1444\r\n[Epoch 95] TestAcc=71.00%  CE=1.1543\r\n[Epoch 96] TestAcc=68.60%  CE=1.1607\r\n[Epoch 97] TestAcc=70.60%  CE=1.1534\r\n[Epoch 98] TestAcc=73.40%  CE=1.1265\r\n[Epoch 99] TestAcc=71.60%  CE=1.1361\r\n[Epoch 100] TestAcc=71.60%  CE=1.1397\r\n\r\n=== NBMF \u5b9f\u9a13 2\/3 ===\r\n[Epoch 01] TestAcc=31.80%  CE=2.2092\r\n[Epoch 02] TestAcc=65.40%  CE=1.5950\r\n[Epoch 03] TestAcc=69.40%  CE=1.3513\r\n[Epoch 04] TestAcc=68.00%  CE=1.2713\r\n[Epoch 05] TestAcc=66.40%  CE=1.2722\r\n[Epoch 06] TestAcc=68.20%  CE=1.2448\r\n[Epoch 07] TestAcc=68.60%  CE=1.1960\r\n[Epoch 08] TestAcc=69.20%  CE=1.1792\r\n[Epoch 09] TestAcc=70.60%  CE=1.1733\r\n[Epoch 10] TestAcc=71.80%  CE=1.1548\r\n[Epoch 11] TestAcc=69.00%  CE=1.1839\r\n[Epoch 12] TestAcc=71.00%  CE=1.1545\r\n[Epoch 13] TestAcc=70.80%  CE=1.1652\r\n[Epoch 14] TestAcc=71.00%  CE=1.1800\r\n[Epoch 15] TestAcc=72.40%  CE=1.1364\r\n[Epoch 16] TestAcc=72.00%  CE=1.1519\r\n[Epoch 17] TestAcc=68.80%  CE=1.1872\r\n[Epoch 18] TestAcc=70.00%  CE=1.1868\r\n[Epoch 19] TestAcc=70.80%  CE=1.1605\r\n[Epoch 20] TestAcc=71.00%  CE=1.1794\r\n[Epoch 21] TestAcc=71.60%  CE=1.1579\r\n[Epoch 22] TestAcc=71.00%  CE=1.1664\r\n[Epoch 23] TestAcc=71.20%  CE=1.1662\r\n[Epoch 24] TestAcc=71.00%  CE=1.1321\r\n[Epoch 25] TestAcc=72.20%  CE=1.1444\r\n[Epoch 26] TestAcc=70.80%  CE=1.1453\r\n[Epoch 27] TestAcc=71.00%  CE=1.1421\r\n[Epoch 28] TestAcc=71.20%  CE=1.1390\r\n[Epoch 29] TestAcc=71.80%  CE=1.1445\r\n[Epoch 30] TestAcc=72.40%  CE=1.1363\r\n[Epoch 31] TestAcc=72.20%  CE=1.1555\r\n[Epoch 32] TestAcc=70.60%  CE=1.1427\r\n[Epoch 33] TestAcc=68.60%  CE=1.1869\r\n[Epoch 34] TestAcc=72.00%  CE=1.1236\r\n[Epoch 35] TestAcc=72.00%  CE=1.1565\r\n[Epoch 36] TestAcc=69.20%  CE=1.1624\r\n[Epoch 37] TestAcc=71.40%  CE=1.1501\r\n[Epoch 38] TestAcc=71.00%  CE=1.1585\r\n[Epoch 39] TestAcc=69.40%  CE=1.1394\r\n[Epoch 40] TestAcc=69.20%  CE=1.1483\r\n[Epoch 41] TestAcc=71.00%  CE=1.1364\r\n[Epoch 42] TestAcc=70.20%  CE=1.1606\r\n[Epoch 43] TestAcc=71.20%  CE=1.1486\r\n[Epoch 44] TestAcc=72.00%  CE=1.1446\r\n[Epoch 45] TestAcc=69.60%  CE=1.1535\r\n[Epoch 46] TestAcc=70.60%  CE=1.1408\r\n[Epoch 47] TestAcc=70.20%  CE=1.1599\r\n[Epoch 48] TestAcc=72.20%  CE=1.1432\r\n[Epoch 49] TestAcc=71.20%  CE=1.1490\r\n[Epoch 50] TestAcc=71.20%  CE=1.1288\r\n[Epoch 51] TestAcc=72.00%  CE=1.1429\r\n[Epoch 52] TestAcc=72.60%  CE=1.1390\r\n[Epoch 53] TestAcc=70.60%  CE=1.1438\r\n[Epoch 54] TestAcc=69.40%  CE=1.1719\r\n[Epoch 55] TestAcc=70.40%  CE=1.1575\r\n[Epoch 56] TestAcc=72.00%  CE=1.1343\r\n[Epoch 57] TestAcc=69.40%  CE=1.1474\r\n[Epoch 58] TestAcc=72.40%  CE=1.1187\r\n[Epoch 59] TestAcc=69.20%  CE=1.1738\r\n[Epoch 60] TestAcc=70.40%  CE=1.1603\r\n[Epoch 61] TestAcc=70.20%  CE=1.1553\r\n[Epoch 62] TestAcc=71.20%  CE=1.1397\r\n[Epoch 63] TestAcc=70.60%  CE=1.1486\r\n[Epoch 64] TestAcc=71.80%  CE=1.1349\r\n[Epoch 65] TestAcc=70.40%  CE=1.1328\r\n[Epoch 66] TestAcc=71.40%  CE=1.1468\r\n[Epoch 67] TestAcc=71.00%  CE=1.1526\r\n[Epoch 68] TestAcc=69.40%  CE=1.1640\r\n[Epoch 69] TestAcc=70.60%  CE=1.1618\r\n[Epoch 70] TestAcc=72.80%  CE=1.1495\r\n[Epoch 71] TestAcc=70.60%  CE=1.1532\r\n[Epoch 72] TestAcc=70.60%  CE=1.1680\r\n[Epoch 73] TestAcc=72.40%  CE=1.1387\r\n[Epoch 74] TestAcc=70.80%  CE=1.1525\r\n[Epoch 75] TestAcc=70.80%  CE=1.1353\r\n[Epoch 76] TestAcc=71.00%  CE=1.1529\r\n[Epoch 77] TestAcc=72.20%  CE=1.1446\r\n[Epoch 78] TestAcc=72.60%  CE=1.1411\r\n[Epoch 79] TestAcc=72.60%  CE=1.1321\r\n[Epoch 80] TestAcc=69.20%  CE=1.1689\r\n[Epoch 81] TestAcc=70.80%  CE=1.1564\r\n[Epoch 82] TestAcc=69.80%  CE=1.1779\r\n[Epoch 83] TestAcc=71.20%  CE=1.1699\r\n[Epoch 84] TestAcc=69.60%  CE=1.1610\r\n[Epoch 85] TestAcc=69.60%  CE=1.1524\r\n[Epoch 86] TestAcc=70.60%  CE=1.1522\r\n[Epoch 87] TestAcc=71.40%  CE=1.1700\r\n[Epoch 88] TestAcc=72.40%  CE=1.1541\r\n[Epoch 89] TestAcc=69.80%  CE=1.1610\r\n[Epoch 90] TestAcc=71.00%  CE=1.1500\r\n[Epoch 91] TestAcc=70.40%  CE=1.1537\r\n[Epoch 92] TestAcc=71.80%  CE=1.1498\r\n[Epoch 93] TestAcc=72.00%  CE=1.1543\r\n[Epoch 94] TestAcc=70.40%  CE=1.1673\r\n[Epoch 95] TestAcc=71.20%  CE=1.1432\r\n[Epoch 96] TestAcc=70.00%  CE=1.1529\r\n[Epoch 97] TestAcc=71.00%  CE=1.1731\r\n[Epoch 98] TestAcc=71.40%  CE=1.1515\r\n[Epoch 99] TestAcc=71.00%  CE=1.1575\r\n[Epoch 100] TestAcc=70.40%  CE=1.1644\r\n\r\n=== NBMF \u5b9f\u9a13 3\/3 ===\r\n[Epoch 01] TestAcc=33.00%  CE=2.1953\r\n[Epoch 02] TestAcc=70.00%  CE=1.6252\r\n[Epoch 03] TestAcc=68.20%  CE=1.4090\r\n[Epoch 04] TestAcc=70.00%  CE=1.3100\r\n[Epoch 05] TestAcc=66.00%  CE=1.3192\r\n[Epoch 06] TestAcc=69.20%  CE=1.2900\r\n[Epoch 07] TestAcc=68.80%  CE=1.2769\r\n[Epoch 08] TestAcc=69.60%  CE=1.2596\r\n[Epoch 09] TestAcc=71.80%  CE=1.2448\r\n[Epoch 10] TestAcc=70.60%  CE=1.2674\r\n[Epoch 11] TestAcc=71.20%  CE=1.2709\r\n[Epoch 12] TestAcc=70.20%  CE=1.2540\r\n[Epoch 13] TestAcc=70.40%  CE=1.2551\r\n[Epoch 14] TestAcc=71.20%  CE=1.2641\r\n[Epoch 15] TestAcc=72.60%  CE=1.2348\r\n[Epoch 16] TestAcc=71.60%  CE=1.2561\r\n[Epoch 17] TestAcc=68.20%  CE=1.2775\r\n[Epoch 18] TestAcc=71.00%  CE=1.2439\r\n[Epoch 19] TestAcc=71.00%  CE=1.2459\r\n[Epoch 20] TestAcc=70.00%  CE=1.2509\r\n[Epoch 21] TestAcc=68.80%  CE=1.2751\r\n[Epoch 22] TestAcc=70.00%  CE=1.2828\r\n[Epoch 23] TestAcc=69.80%  CE=1.2526\r\n[Epoch 24] TestAcc=70.40%  CE=1.2449\r\n[Epoch 25] TestAcc=72.40%  CE=1.2370\r\n[Epoch 26] TestAcc=70.60%  CE=1.2540\r\n[Epoch 27] TestAcc=71.80%  CE=1.2561\r\n[Epoch 28] TestAcc=69.40%  CE=1.2918\r\n[Epoch 29] TestAcc=71.20%  CE=1.2554\r\n[Epoch 30] TestAcc=71.20%  CE=1.2518\r\n[Epoch 31] TestAcc=71.40%  CE=1.2559\r\n[Epoch 32] TestAcc=70.20%  CE=1.2719\r\n[Epoch 33] TestAcc=70.40%  CE=1.2587\r\n[Epoch 34] TestAcc=70.40%  CE=1.2505\r\n[Epoch 35] TestAcc=71.20%  CE=1.2584\r\n[Epoch 36] TestAcc=70.00%  CE=1.2434\r\n[Epoch 37] TestAcc=71.80%  CE=1.2551\r\n[Epoch 38] TestAcc=70.40%  CE=1.2571\r\n[Epoch 39] TestAcc=69.40%  CE=1.2872\r\n[Epoch 40] TestAcc=70.40%  CE=1.2501\r\n[Epoch 41] TestAcc=67.80%  CE=1.2883\r\n[Epoch 42] TestAcc=72.00%  CE=1.2382\r\n[Epoch 43] TestAcc=70.80%  CE=1.2543\r\n[Epoch 44] TestAcc=69.40%  CE=1.2393\r\n[Epoch 45] TestAcc=71.00%  CE=1.2334\r\n[Epoch 46] TestAcc=69.80%  CE=1.2566\r\n[Epoch 47] TestAcc=69.00%  CE=1.2686\r\n[Epoch 48] TestAcc=68.00%  CE=1.2761\r\n[Epoch 49] TestAcc=71.20%  CE=1.2381\r\n[Epoch 50] TestAcc=71.20%  CE=1.2498\r\n[Epoch 51] TestAcc=74.00%  CE=1.2131\r\n[Epoch 52] TestAcc=68.80%  CE=1.2677\r\n[Epoch 53] TestAcc=72.00%  CE=1.2469\r\n[Epoch 54] TestAcc=71.60%  CE=1.2377\r\n[Epoch 55] TestAcc=68.20%  CE=1.2533\r\n[Epoch 56] TestAcc=70.40%  CE=1.2671\r\n[Epoch 57] TestAcc=71.00%  CE=1.2565\r\n[Epoch 58] TestAcc=70.80%  CE=1.2545\r\n[Epoch 59] TestAcc=69.40%  CE=1.2441\r\n[Epoch 60] TestAcc=69.60%  CE=1.2542\r\n[Epoch 61] TestAcc=71.00%  CE=1.2439\r\n[Epoch 62] TestAcc=71.20%  CE=1.2277\r\n[Epoch 63] TestAcc=72.60%  CE=1.2147\r\n[Epoch 64] TestAcc=71.80%  CE=1.2430\r\n[Epoch 65] TestAcc=70.80%  CE=1.2659\r\n[Epoch 66] TestAcc=70.60%  CE=1.2561\r\n[Epoch 67] TestAcc=73.20%  CE=1.2252\r\n[Epoch 68] TestAcc=72.80%  CE=1.2408\r\n[Epoch 69] TestAcc=69.20%  CE=1.2425\r\n[Epoch 70] TestAcc=68.40%  CE=1.2747\r\n[Epoch 71] TestAcc=69.40%  CE=1.2675\r\n[Epoch 72] TestAcc=69.60%  CE=1.2309\r\n[Epoch 73] TestAcc=72.80%  CE=1.1989\r\n[Epoch 74] TestAcc=72.40%  CE=1.1864\r\n[Epoch 75] TestAcc=74.00%  CE=1.1654\r\n[Epoch 76] TestAcc=73.40%  CE=1.1872\r\n[Epoch 77] TestAcc=72.20%  CE=1.1872\r\n[Epoch 78] TestAcc=70.20%  CE=1.2198\r\n[Epoch 79] TestAcc=69.80%  CE=1.2250\r\n[Epoch 80] TestAcc=73.00%  CE=1.1992\r\n[Epoch 81] TestAcc=70.00%  CE=1.2213\r\n[Epoch 82] TestAcc=70.80%  CE=1.2078\r\n[Epoch 83] TestAcc=71.60%  CE=1.2019\r\n[Epoch 84] TestAcc=70.80%  CE=1.2283\r\n[Epoch 85] TestAcc=72.20%  CE=1.1838\r\n[Epoch 86] TestAcc=72.60%  CE=1.2193\r\n[Epoch 87] TestAcc=72.00%  CE=1.2178\r\n[Epoch 88] TestAcc=69.20%  CE=1.2068\r\n[Epoch 89] TestAcc=73.00%  CE=1.2022\r\n[Epoch 90] TestAcc=74.00%  CE=1.1804\r\n[Epoch 91] TestAcc=71.20%  CE=1.2008\r\n[Epoch 92] TestAcc=69.00%  CE=1.1980\r\n[Epoch 93] TestAcc=72.40%  CE=1.1960\r\n[Epoch 94] TestAcc=71.80%  CE=1.2141\r\n[Epoch 95] TestAcc=72.60%  CE=1.1906\r\n[Epoch 96] TestAcc=72.00%  CE=1.2227\r\n[Epoch 97] TestAcc=70.40%  CE=1.2087\r\n[Epoch 98] TestAcc=70.80%  CE=1.2104\r\n[Epoch 99] TestAcc=72.40%  CE=1.1876\r\n[Epoch 100] TestAcc=70.20%  CE=1.2316\r\n<\/pre>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<\/details>\n<div class=\"cell border-box-sizing text_cell rendered\">\n<div class=\"inner_cell\">\n<div class=\"text_cell_render border-box-sizing rendered_html\">\n<p>FCNN\u3068NBMF\u305d\u308c\u305e\u308c\u306e\u5e73\u5747\u7cbe\u5ea6\u3068\u4ea4\u5dee\u30a8\u30f3\u30c8\u30ed\u30d4\u30fc\u8aa4\u5dee\u3092\u8868\u793a\u3057\u307e\u3059\u3002<\/p>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing code_cell rendered\">\n<div class=\"input\">\n<div class=\"inner_cell\">\n<div class=\"input_area\">\n<div class=\"highlight hl-python\">\n<pre><span><\/span><span class=\"nb\">print<\/span><span class=\"p\">(<\/span><span class=\"s2\">\"<\/span><span class=\"se\">\\n<\/span><span class=\"s2\">\u5e73\u5747\u7cbe\u5ea6\uff08FCNN\uff09:\"<\/span><span class=\"p\">,<\/span> <span class=\"n\">acc_mean_fcnn<\/span><span class=\"p\">[<\/span><span class=\"o\">-<\/span><span class=\"mi\">1<\/span><span class=\"p\">]<\/span><span class=\"o\">*<\/span><span class=\"mi\">100<\/span><span class=\"p\">,<\/span> <span class=\"s2\">\"%\"<\/span><span class=\"p\">)<\/span>\r\n<span class=\"nb\">print<\/span><span class=\"p\">(<\/span><span class=\"s2\">\"\u5e73\u5747\u4ea4\u5dee\u30a8\u30f3\u30c8\u30ed\u30d4\u30fc\u8aa4\u5dee\uff08FCNN\uff09:\"<\/span><span class=\"p\">,<\/span> <span class=\"n\">ce_mean_fcnn<\/span><span class=\"p\">[<\/span><span class=\"o\">-<\/span><span class=\"mi\">1<\/span><span class=\"p\">])<\/span>\r\n<span class=\"nb\">print<\/span><span class=\"p\">(<\/span><span class=\"s2\">\"\u5e73\u5747\u7cbe\u5ea6\uff08NBMF\uff09:\"<\/span><span class=\"p\">,<\/span> <span class=\"n\">acc_mean_nbmf<\/span><span class=\"p\">[<\/span><span class=\"o\">-<\/span><span class=\"mi\">1<\/span><span class=\"p\">]<\/span><span class=\"o\">*<\/span><span class=\"mi\">100<\/span><span class=\"p\">,<\/span> <span class=\"s2\">\"%\"<\/span><span class=\"p\">)<\/span>\r\n<span class=\"nb\">print<\/span><span class=\"p\">(<\/span><span class=\"s2\">\"\u5e73\u5747\u4ea4\u5dee\u30a8\u30f3\u30c8\u30ed\u30d4\u30fc\u8aa4\u5dee\uff08NBMF\uff09:\"<\/span><span class=\"p\">,<\/span> <span class=\"n\">ce_mean_nbmf<\/span><span class=\"p\">[<\/span><span class=\"o\">-<\/span><span class=\"mi\">1<\/span><span class=\"p\">])<\/span>\r\n<\/pre>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"output_wrapper\">\n<div class=\"output\">\n<div class=\"output_area\">\n<div class=\"prompt\"><\/div>\n<div class=\"output_subarea output_stream output_stdout output_text\">\n<pre>\u5e73\u5747\u7cbe\u5ea6\uff08FCNN\uff09: 84.39999999999999 %\r\n\u5e73\u5747\u4ea4\u5dee\u30a8\u30f3\u30c8\u30ed\u30d4\u30fc\u8aa4\u5dee\uff08FCNN\uff09: 0.5280500643255045\r\n\u5e73\u5747\u7cbe\u5ea6\uff08NBMF\uff09: 70.73333333333332 %\r\n\u5e73\u5747\u4ea4\u5dee\u30a8\u30f3\u30c8\u30ed\u30d4\u30fc\u8aa4\u5dee\uff08NBMF\uff09: 1.1785668295091172\r\n<\/pre>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing text_cell rendered\">\n<div class=\"inner_cell\">\n<div class=\"text_cell_render border-box-sizing rendered_html\">\n<p>\u5b66\u7fd2\u66f2\u7dda\u3092\u63cf\u753b\u3057\u307e\u3059\u3002<\/p>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing code_cell rendered\">\n<div class=\"input\">\n<div class=\"inner_cell\">\n<div class=\"input_area\">\n<div class=\"highlight hl-python\">\n<pre><span><\/span><span class=\"n\">plot_learning_curves<\/span><span class=\"p\">(<\/span>\r\n    <span class=\"n\">fcnn_epochs<\/span><span class=\"p\">,<\/span> <span class=\"n\">nbmf_epochs<\/span><span class=\"p\">,<\/span>\r\n    <span class=\"n\">acc_mean_fcnn<\/span><span class=\"p\">,<\/span> <span class=\"n\">acc_std_fcnn<\/span><span class=\"p\">,<\/span> <span class=\"n\">ce_mean_fcnn<\/span><span class=\"p\">,<\/span> <span class=\"n\">ce_std_fcnn<\/span><span class=\"p\">,<\/span>\r\n    <span class=\"n\">acc_mean_nbmf<\/span><span class=\"p\">,<\/span> <span class=\"n\">acc_std_nbmf<\/span><span class=\"p\">,<\/span> <span class=\"n\">ce_mean_nbmf<\/span><span class=\"p\">,<\/span> <span class=\"n\">ce_std_nbmf<\/span>\r\n<span class=\"p\">)<\/span>\r\n<\/pre>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"output_wrapper\">\n<div class=\"output\">\n<div class=\"output_area\">\n<div class=\"prompt\"><\/div>\n<div class=\"output_png output_subarea\"><img decoding=\"async\" alt=\"No description has been provided for this image\" src=\"data:image\/png;base64,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\" \/><\/div>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing text_cell rendered\">\n<div class=\"inner_cell\">\n<div class=\"text_cell_render border-box-sizing rendered_html\">\n<h3><span class=\"ez-toc-section\" id=\"W%E3%81%AE%E4%B8%AD%E8%BA%AB%E3%81%AE%E5%8F%AF%E8%A6%96%E5%8C%96\"><\/span>$W$\u306e\u4e2d\u8eab\u306e\u53ef\u8996\u5316<span class=\"ez-toc-section-end\"><\/span><\/h3>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing text_cell rendered\">\n<div class=\"inner_cell\">\n<div class=\"text_cell_render border-box-sizing rendered_html\">\n<p>\u30a8\u30dd\u30c3\u30af\u6570\u306b\u5bfe\u3059\u308b\u7cbe\u5ea6\u3068\u4ea4\u5dee\u30a8\u30f3\u30c8\u30ed\u30d4\u30fc\u8aa4\u5dee\u3092\u6bd4\u8f03\u3057\u305f\u5b9f\u9a13\u306b\u5bfe\u3057\u3066\u3001$W$\u306e\u4e2d\u8eab\u3068\u7d50\u679c\u3092\u53ef\u8996\u5316\u3057\u3066\u307f\u307e\u3059\u3002<br \/>\n\u4ee5\u4e0b\u3067\u306f\u3001\u307e\u305a$W$\u306e\u4e2d\u8eab\u306e\u753b\u50cf\u3092\u8868\u3059\u90e8\u5206\u3092\u8868\u793a\u3059\u308b\u95a2\u6570\u3001\u305d\u306e\u753b\u50cf\u306e\u30af\u30e9\u30b9\u60c5\u5831\u3092\u8868\u3059\u30d2\u30b9\u30c8\u30b0\u30e9\u30e0\u3092\u63cf\u753b\u3059\u308b\u95a2\u6570\u3001\u305d\u306e\u753b\u50cf\u304c\u305d\u308c\u305e\u308c\u306e\u6570\u5b57\u3092\u518d\u69cb\u6210\u3059\u308b\u969b\u306b\u4f55\u56de\u7528\u3044\u3089\u308c\u305f\u304b\u3092\u8868\u3059\u30d2\u30b9\u30c8\u30b0\u30e9\u30e0\u3092\u63cf\u753b\u3059\u308b\u95a2\u6570\u3092\u4f5c\u6210\u3057\u307e\u3059\u3002<\/p>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing code_cell rendered\">\n<div class=\"input\">\n<div class=\"inner_cell\">\n<div class=\"input_area\">\n<div class=\"highlight hl-python\">\n<pre><span><\/span><span class=\"k\">def<\/span><span class=\"w\"> <\/span><span class=\"nf\">show_W_column<\/span><span class=\"p\">(<\/span><span class=\"n\">W1<\/span><span class=\"p\">,<\/span> <span class=\"n\">idx<\/span><span class=\"p\">):<\/span>\r\n<span class=\"w\">    <\/span><span class=\"sd\">\"\"\"<\/span>\r\n<span class=\"sd\">    W1\uff08\u753b\u50cf\u90e8\u5206\u306e\u57fa\u5e95\u884c\u5217\uff09\u306e\u7279\u5b9a\u306e\u5217\u3092\u753b\u50cf\u3068\u3057\u3066\u8868\u793a\u3059\u308b<\/span>\r\n<span class=\"sd\">    &lt;\u5165\u529b&gt;<\/span>\r\n<span class=\"sd\">    W1\uff1a(784, k), \u753b\u50cf\u6210\u5206\u306e\u57fa\u5e95\u884c\u5217<\/span>\r\n<span class=\"sd\">    idx\uff1a(int), \u8868\u793a\u3059\u308b\u57fa\u5e95\u306e\u30a4\u30f3\u30c7\u30c3\u30af\u30b9<\/span>\r\n<span class=\"sd\">    &lt;\u51fa\u529b&gt;<\/span>\r\n<span class=\"sd\">    \u306a\u3057<\/span>\r\n<span class=\"sd\">    \"\"\"<\/span>\r\n    <span class=\"n\">img<\/span> <span class=\"o\">=<\/span> <span class=\"n\">W1<\/span><span class=\"p\">[:,<\/span> <span class=\"n\">idx<\/span><span class=\"p\">]<\/span><span class=\"o\">.<\/span><span class=\"n\">reshape<\/span><span class=\"p\">(<\/span><span class=\"mi\">28<\/span><span class=\"p\">,<\/span> <span class=\"mi\">28<\/span><span class=\"p\">)<\/span>\r\n    <span class=\"n\">plt<\/span><span class=\"o\">.<\/span><span class=\"n\">imshow<\/span><span class=\"p\">(<\/span><span class=\"n\">img<\/span><span class=\"p\">,<\/span> <span class=\"n\">cmap<\/span><span class=\"o\">=<\/span><span class=\"s1\">'gray'<\/span><span class=\"p\">)<\/span>\r\n    <span class=\"n\">plt<\/span><span class=\"o\">.<\/span><span class=\"n\">axis<\/span><span class=\"p\">(<\/span><span class=\"s1\">'off'<\/span><span class=\"p\">)<\/span>\r\n\r\n<span class=\"k\">def<\/span><span class=\"w\"> <\/span><span class=\"nf\">compute_usage_histogram<\/span><span class=\"p\">(<\/span><span class=\"n\">H<\/span><span class=\"p\">,<\/span> <span class=\"n\">y_train<\/span><span class=\"p\">,<\/span> <span class=\"n\">k<\/span><span class=\"p\">):<\/span>\r\n<span class=\"w\">    <\/span><span class=\"sd\">\"\"\"<\/span>\r\n<span class=\"sd\">    \u5404\u57fa\u5e95\u30d9\u30af\u30c8\u30eb\u304c\u3069\u306e\u30af\u30e9\u30b9\uff08\u6570\u5b57\u30e9\u30d9\u30eb\uff09\u306e\u30c7\u30fc\u30bf\u518d\u69cb\u6210\u306b\u4f7f\u7528\u3055\u308c\u305f\u304b\u3092\u96c6\u8a08\u3059\u308b<\/span>\r\n<span class=\"sd\">    &lt;\u5165\u529b&gt;<\/span>\r\n<span class=\"sd\">    H\uff1a(k, n_samples), \u4fc2\u6570\u884c\u5217<\/span>\r\n<span class=\"sd\">    y_train\uff1a(n_samples,), \u6559\u5e2b\u30c7\u30fc\u30bf\u306e\u6b63\u89e3\u30e9\u30d9\u30eb<\/span>\r\n<span class=\"sd\">    k\uff1a(int), \u57fa\u5e95\u30d9\u30af\u30c8\u30eb\u306e\u6570<\/span>\r\n<span class=\"sd\">    &lt;\u51fa\u529b&gt;<\/span>\r\n<span class=\"sd\">    usage\uff1a(10, k), \u30af\u30e9\u30b9\u3054\u3068\u306e\u57fa\u5e95\u4f7f\u7528\u56de\u6570\u3092\u683c\u7d0d\u3057\u305f\u884c\u5217<\/span>\r\n<span class=\"sd\">    \"\"\"<\/span>\r\n    <span class=\"n\">usage<\/span> <span class=\"o\">=<\/span> <span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">zeros<\/span><span class=\"p\">((<\/span><span class=\"mi\">10<\/span><span class=\"p\">,<\/span> <span class=\"n\">k<\/span><span class=\"p\">),<\/span> <span class=\"n\">dtype<\/span><span class=\"o\">=<\/span><span class=\"nb\">int<\/span><span class=\"p\">)<\/span>\r\n\r\n    <span class=\"k\">for<\/span> <span class=\"n\">img_idx<\/span> <span class=\"ow\">in<\/span> <span class=\"nb\">range<\/span><span class=\"p\">(<\/span><span class=\"n\">H<\/span><span class=\"o\">.<\/span><span class=\"n\">shape<\/span><span class=\"p\">[<\/span><span class=\"mi\">1<\/span><span class=\"p\">]):<\/span>\r\n        <span class=\"n\">active<\/span> <span class=\"o\">=<\/span> <span class=\"n\">H<\/span><span class=\"p\">[:,<\/span> <span class=\"n\">img_idx<\/span><span class=\"p\">]<\/span> <span class=\"o\">&gt;<\/span> <span class=\"mi\">0<\/span>\r\n        <span class=\"n\">label<\/span> <span class=\"o\">=<\/span> <span class=\"n\">y_train<\/span><span class=\"p\">[<\/span><span class=\"n\">img_idx<\/span><span class=\"p\">]<\/span>\r\n        <span class=\"n\">usage<\/span><span class=\"p\">[<\/span><span class=\"n\">label<\/span><span class=\"p\">,<\/span> <span class=\"n\">active<\/span><span class=\"p\">]<\/span> <span class=\"o\">+=<\/span> <span class=\"mi\">1<\/span>\r\n\r\n    <span class=\"k\">return<\/span> <span class=\"n\">usage<\/span>\r\n\r\n<span class=\"k\">def<\/span><span class=\"w\"> <\/span><span class=\"nf\">plot_usage_histogram<\/span><span class=\"p\">(<\/span><span class=\"n\">usage<\/span><span class=\"p\">,<\/span> <span class=\"n\">idx<\/span><span class=\"p\">):<\/span>\r\n<span class=\"w\">    <\/span><span class=\"sd\">\"\"\"<\/span>\r\n<span class=\"sd\">    \u7279\u5b9a\u306e\u57fa\u5e95\u304c\u5404\u30af\u30e9\u30b9\uff080-9\uff09\u3067\u3069\u308c\u304f\u3089\u3044\u306e\u983b\u5ea6\u3067\u4f7f\u7528\u3055\u308c\u305f\u304b\u3092\u68d2\u30b0\u30e9\u30d5\u3067\u8868\u793a\u3059\u308b<\/span>\r\n<span class=\"sd\">    &lt;\u5165\u529b&gt;<\/span>\r\n<span class=\"sd\">    usage\uff1a(10, k), compute_usage_histogram\u3067\u8a08\u7b97\u3055\u308c\u305f\u4f7f\u7528\u983b\u5ea6\u884c\u5217<\/span>\r\n<span class=\"sd\">    idx\uff1a(int), \u8868\u793a\u5bfe\u8c61\u306e\u57fa\u5e95\u30a4\u30f3\u30c7\u30c3\u30af\u30b9<\/span>\r\n<span class=\"sd\">    &lt;\u51fa\u529b&gt;<\/span>\r\n<span class=\"sd\">    \u306a\u3057<\/span>\r\n<span class=\"sd\">    \"\"\"<\/span>\r\n    <span class=\"n\">plt<\/span><span class=\"o\">.<\/span><span class=\"n\">bar<\/span><span class=\"p\">(<\/span><span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">arange<\/span><span class=\"p\">(<\/span><span class=\"mi\">10<\/span><span class=\"p\">),<\/span> <span class=\"n\">usage<\/span><span class=\"p\">[:,<\/span> <span class=\"n\">idx<\/span><span class=\"p\">])<\/span>\r\n    <span class=\"n\">plt<\/span><span class=\"o\">.<\/span><span class=\"n\">xlabel<\/span><span class=\"p\">(<\/span><span class=\"s2\">\"Label\"<\/span><span class=\"p\">)<\/span>\r\n    <span class=\"n\">plt<\/span><span class=\"o\">.<\/span><span class=\"n\">ylabel<\/span><span class=\"p\">(<\/span><span class=\"s2\">\"Frequency\"<\/span><span class=\"p\">)<\/span>\r\n    <span class=\"n\">plt<\/span><span class=\"o\">.<\/span><span class=\"n\">title<\/span><span class=\"p\">(<\/span><span class=\"sa\">f<\/span><span class=\"s2\">\"Usage histogram for W column <\/span><span class=\"si\">{<\/span><span class=\"n\">idx<\/span><span class=\"si\">}<\/span><span class=\"s2\">\"<\/span><span class=\"p\">)<\/span>\r\n\r\n<span class=\"k\">def<\/span><span class=\"w\"> <\/span><span class=\"nf\">plot_W2_bar<\/span><span class=\"p\">(<\/span><span class=\"n\">W2<\/span><span class=\"p\">,<\/span> <span class=\"n\">idx<\/span><span class=\"p\">):<\/span>\r\n<span class=\"w\">    <\/span><span class=\"sd\">\"\"\"<\/span>\r\n<span class=\"sd\">    \u30e9\u30d9\u30eb\u60c5\u5831\u90e8\u5206\uff08W2\uff09\u306e\u91cd\u307f\u306e\u68d2\u30b0\u30e9\u30d5\u3092\u8868\u793a\u3059\u308b<\/span>\r\n<span class=\"sd\">    &lt;\u5165\u529b&gt;<\/span>\r\n<span class=\"sd\">    W2\uff1a(10, k), \u30e9\u30d9\u30eb\u6210\u5206\u306e\u57fa\u5e95\u884c\u5217<\/span>\r\n<span class=\"sd\">    idx\uff1a(int), \u8868\u793a\u5bfe\u8c61\u306e\u57fa\u5e95\u30a4\u30f3\u30c7\u30c3\u30af\u30b9<\/span>\r\n<span class=\"sd\">    &lt;\u51fa\u529b&gt;<\/span>\r\n<span class=\"sd\">    \u306a\u3057<\/span>\r\n<span class=\"sd\">    \"\"\"<\/span>\r\n    <span class=\"n\">plt<\/span><span class=\"o\">.<\/span><span class=\"n\">bar<\/span><span class=\"p\">(<\/span><span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">arange<\/span><span class=\"p\">(<\/span><span class=\"mi\">10<\/span><span class=\"p\">),<\/span> <span class=\"n\">W2<\/span><span class=\"p\">[:,<\/span> <span class=\"n\">idx<\/span><span class=\"p\">])<\/span>\r\n    <span class=\"n\">plt<\/span><span class=\"o\">.<\/span><span class=\"n\">xlabel<\/span><span class=\"p\">(<\/span><span class=\"s2\">\"Label index\"<\/span><span class=\"p\">)<\/span>\r\n    <span class=\"n\">plt<\/span><span class=\"o\">.<\/span><span class=\"n\">ylabel<\/span><span class=\"p\">(<\/span><span class=\"s2\">\"Value\"<\/span><span class=\"p\">)<\/span>\r\n    <span class=\"n\">plt<\/span><span class=\"o\">.<\/span><span class=\"n\">title<\/span><span class=\"p\">(<\/span><span class=\"sa\">f<\/span><span class=\"s2\">\"W2 contributions for column <\/span><span class=\"si\">{<\/span><span class=\"n\">idx<\/span><span class=\"si\">}<\/span><span class=\"s2\">\"<\/span><span class=\"p\">)<\/span>\r\n\r\n<span class=\"k\">def<\/span><span class=\"w\"> <\/span><span class=\"nf\">plot_W_feature_analysis<\/span><span class=\"p\">(<\/span><span class=\"n\">W<\/span><span class=\"p\">,<\/span> <span class=\"n\">H<\/span><span class=\"p\">,<\/span> <span class=\"n\">y_train<\/span><span class=\"p\">,<\/span> <span class=\"n\">idx<\/span><span class=\"p\">):<\/span>\r\n<span class=\"w\">    <\/span><span class=\"sd\">\"\"\"<\/span>\r\n<span class=\"sd\">    \u7279\u5b9a\u306e\u57fa\u5e95\u306b\u3064\u3044\u3066\u3001\u57fa\u5e95\u753b\u50cf\u30fb\u30af\u30e9\u30b9\u4f7f\u7528\u983b\u5ea6\u30fb\u30e9\u30d9\u30eb\u5bc4\u4e0e\uff08W2\uff09\u3092\u307e\u3068\u3081\u3066\u53ef\u8996\u5316\u3059\u308b<\/span>\r\n<span class=\"sd\">    &lt;\u5165\u529b&gt;<\/span>\r\n<span class=\"sd\">    W\uff1a(794, k), \u753b\u50cf\u6210\u5206\u3068\u30e9\u30d9\u30eb\u6210\u5206\u304c\u7d50\u5408\u3055\u308c\u305f\u57fa\u5e95\u884c\u5217<\/span>\r\n<span class=\"sd\">    H\uff1a(k, n_samples), \u4fc2\u6570\u884c\u5217<\/span>\r\n<span class=\"sd\">    y_train\uff1a(n_samples,), \u6559\u5e2b\u30c7\u30fc\u30bf\u306e\u6b63\u89e3\u30e9\u30d9\u30eb<\/span>\r\n<span class=\"sd\">    idx\uff1a(int), \u5206\u6790\u5bfe\u8c61\u306e\u57fa\u5e95\u30a4\u30f3\u30c7\u30c3\u30af\u30b9<\/span>\r\n<span class=\"sd\">    &lt;\u51fa\u529b&gt;<\/span>\r\n<span class=\"sd\">    \u306a\u3057<\/span>\r\n<span class=\"sd\">    \"\"\"<\/span>\r\n    <span class=\"n\">n_img<\/span> <span class=\"o\">=<\/span> <span class=\"mi\">784<\/span>\r\n    <span class=\"n\">W1<\/span> <span class=\"o\">=<\/span> <span class=\"n\">W<\/span><span class=\"p\">[:<\/span><span class=\"n\">n_img<\/span><span class=\"p\">,<\/span> <span class=\"p\">:]<\/span>\r\n    <span class=\"n\">W2<\/span> <span class=\"o\">=<\/span> <span class=\"n\">W<\/span><span class=\"p\">[<\/span><span class=\"n\">n_img<\/span><span class=\"p\">:,<\/span> <span class=\"p\">:]<\/span>\r\n\r\n    <span class=\"n\">usage<\/span> <span class=\"o\">=<\/span> <span class=\"n\">compute_usage_histogram<\/span><span class=\"p\">(<\/span><span class=\"n\">H<\/span><span class=\"p\">,<\/span> <span class=\"n\">y_train<\/span><span class=\"p\">,<\/span> <span class=\"n\">W<\/span><span class=\"o\">.<\/span><span class=\"n\">shape<\/span><span class=\"p\">[<\/span><span class=\"mi\">1<\/span><span class=\"p\">])<\/span>\r\n\r\n    <span class=\"n\">fig<\/span><span class=\"p\">,<\/span> <span class=\"n\">axes<\/span> <span class=\"o\">=<\/span> <span class=\"n\">plt<\/span><span class=\"o\">.<\/span><span class=\"n\">subplots<\/span><span class=\"p\">(<\/span><span class=\"mi\">1<\/span><span class=\"p\">,<\/span> <span class=\"mi\">3<\/span><span class=\"p\">,<\/span> <span class=\"n\">figsize<\/span><span class=\"o\">=<\/span><span class=\"p\">(<\/span><span class=\"mi\">12<\/span><span class=\"p\">,<\/span> <span class=\"mi\">3<\/span><span class=\"p\">))<\/span>\r\n\r\n    <span class=\"n\">axes<\/span><span class=\"p\">[<\/span><span class=\"mi\">0<\/span><span class=\"p\">]<\/span><span class=\"o\">.<\/span><span class=\"n\">imshow<\/span><span class=\"p\">(<\/span><span class=\"n\">W1<\/span><span class=\"p\">[:,<\/span> <span class=\"n\">idx<\/span><span class=\"p\">]<\/span><span class=\"o\">.<\/span><span class=\"n\">reshape<\/span><span class=\"p\">(<\/span><span class=\"mi\">28<\/span><span class=\"p\">,<\/span> <span class=\"mi\">28<\/span><span class=\"p\">),<\/span> <span class=\"n\">cmap<\/span><span class=\"o\">=<\/span><span class=\"s1\">'gray'<\/span><span class=\"p\">)<\/span>\r\n    <span class=\"n\">axes<\/span><span class=\"p\">[<\/span><span class=\"mi\">0<\/span><span class=\"p\">]<\/span><span class=\"o\">.<\/span><span class=\"n\">set_title<\/span><span class=\"p\">(<\/span><span class=\"sa\">f<\/span><span class=\"s2\">\"W1 column <\/span><span class=\"si\">{<\/span><span class=\"n\">idx<\/span><span class=\"si\">}<\/span><span class=\"s2\">\"<\/span><span class=\"p\">)<\/span>\r\n    <span class=\"n\">axes<\/span><span class=\"p\">[<\/span><span class=\"mi\">0<\/span><span class=\"p\">]<\/span><span class=\"o\">.<\/span><span class=\"n\">axis<\/span><span class=\"p\">(<\/span><span class=\"s1\">'off'<\/span><span class=\"p\">)<\/span>\r\n\r\n    <span class=\"n\">axes<\/span><span class=\"p\">[<\/span><span class=\"mi\">1<\/span><span class=\"p\">]<\/span><span class=\"o\">.<\/span><span class=\"n\">bar<\/span><span class=\"p\">(<\/span><span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">arange<\/span><span class=\"p\">(<\/span><span class=\"mi\">10<\/span><span class=\"p\">),<\/span> <span class=\"n\">usage<\/span><span class=\"p\">[:,<\/span> <span class=\"n\">idx<\/span><span class=\"p\">])<\/span>\r\n    <span class=\"n\">axes<\/span><span class=\"p\">[<\/span><span class=\"mi\">1<\/span><span class=\"p\">]<\/span><span class=\"o\">.<\/span><span class=\"n\">set_title<\/span><span class=\"p\">(<\/span><span class=\"s2\">\"Usage bar graph\"<\/span><span class=\"p\">)<\/span>\r\n    <span class=\"n\">axes<\/span><span class=\"p\">[<\/span><span class=\"mi\">1<\/span><span class=\"p\">]<\/span><span class=\"o\">.<\/span><span class=\"n\">set_xlabel<\/span><span class=\"p\">(<\/span><span class=\"s2\">\"Label\"<\/span><span class=\"p\">)<\/span>\r\n\r\n    <span class=\"n\">axes<\/span><span class=\"p\">[<\/span><span class=\"mi\">2<\/span><span class=\"p\">]<\/span><span class=\"o\">.<\/span><span class=\"n\">bar<\/span><span class=\"p\">(<\/span><span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">arange<\/span><span class=\"p\">(<\/span><span class=\"mi\">10<\/span><span class=\"p\">),<\/span> <span class=\"n\">W2<\/span><span class=\"p\">[:,<\/span> <span class=\"n\">idx<\/span><span class=\"p\">])<\/span>\r\n    <span class=\"n\">axes<\/span><span class=\"p\">[<\/span><span class=\"mi\">2<\/span><span class=\"p\">]<\/span><span class=\"o\">.<\/span><span class=\"n\">set_title<\/span><span class=\"p\">(<\/span><span class=\"s2\">\"W2 contribution\"<\/span><span class=\"p\">)<\/span>\r\n    <span class=\"n\">axes<\/span><span class=\"p\">[<\/span><span class=\"mi\">2<\/span><span class=\"p\">]<\/span><span class=\"o\">.<\/span><span class=\"n\">set_xlabel<\/span><span class=\"p\">(<\/span><span class=\"s2\">\"Label\"<\/span><span class=\"p\">)<\/span>\r\n\r\n    <span class=\"n\">plt<\/span><span class=\"o\">.<\/span><span class=\"n\">tight_layout<\/span><span class=\"p\">()<\/span>\r\n    <span class=\"n\">plt<\/span><span class=\"o\">.<\/span><span class=\"n\">show<\/span><span class=\"p\">()<\/span>\r\n<\/pre>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing text_cell rendered\">\n<div class=\"inner_cell\">\n<div class=\"text_cell_render border-box-sizing rendered_html\">\n<p>\u4f5c\u6210\u3057\u305f\u95a2\u6570\u3092\u7528\u3044\u3066$W_1$\u306e\u7279\u5b9a\u306e1\u5217\u304c\u8868\u3059\u753b\u50cf\u3068\u305d\u306e\u5217\u304c\u305d\u308c\u305e\u308c\u306e\u6570\u5b57\u3092\u518d\u69cb\u6210\u3059\u308b\u306e\u306b\u7528\u3044\u3089\u308c\u305f\u56de\u6570\u3001\u304a\u3088\u3073$W_2$\u306e\u540c\u3058\u5217\u306e\u60c5\u5831\u3092\u8868\u3059\u68d2\u30b0\u30e9\u30d5\u3092\u63cf\u753b\u3057\u307e\u3059\u3002<\/p>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing code_cell rendered\">\n<div class=\"input\">\n<div class=\"inner_cell\">\n<div class=\"input_area\">\n<div class=\"highlight hl-python\">\n<pre><span><\/span><span class=\"nb\">print<\/span><span class=\"p\">(<\/span><span class=\"s2\">\"=== W \u306e\u57fa\u5e95\u53ef\u8996\u5316\u3068\u4f7f\u7528\u983b\u5ea6 ===\"<\/span><span class=\"p\">)<\/span>\r\n\r\n<span class=\"n\">num<\/span> <span class=\"o\">=<\/span> <span class=\"mi\">5<\/span>  <span class=\"c1\"># \u8868\u793a\u3055\u305b\u308b\u500b\u6570<\/span>\r\n<span class=\"n\">feature_ids<\/span> <span class=\"o\">=<\/span> <span class=\"nb\">list<\/span><span class=\"p\">(<\/span><span class=\"nb\">range<\/span><span class=\"p\">(<\/span><span class=\"nb\">min<\/span><span class=\"p\">(<\/span><span class=\"n\">k<\/span><span class=\"p\">,<\/span> <span class=\"n\">num<\/span><span class=\"p\">)))<\/span>\r\n\r\n<span class=\"k\">for<\/span> <span class=\"n\">idx<\/span> <span class=\"ow\">in<\/span> <span class=\"n\">feature_ids<\/span><span class=\"p\">:<\/span>\r\n    <span class=\"n\">plot_W_feature_analysis<\/span><span class=\"p\">(<\/span><span class=\"n\">W_final<\/span><span class=\"p\">,<\/span> <span class=\"n\">H_final<\/span><span class=\"p\">,<\/span> <span class=\"n\">y_train<\/span><span class=\"p\">,<\/span> <span class=\"n\">idx<\/span><span class=\"p\">)<\/span>\r\n<\/pre>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"output_wrapper\">\n<div class=\"output\">\n<div class=\"output_area\">\n<div class=\"prompt\"><\/div>\n<div class=\"output_subarea output_stream output_stdout output_text\">\n<pre>=== W \u306e\u57fa\u5e95\u53ef\u8996\u5316\u3068\u4f7f\u7528\u983b\u5ea6 ===\r\n<\/pre>\n<\/div>\n<\/div>\n<div class=\"output_area\">\n<div class=\"prompt\"><\/div>\n<div class=\"output_png output_subarea\"><img decoding=\"async\" alt=\"No description has been provided for this image\" src=\"data:image\/png;base64,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\" \/><\/div>\n<\/div>\n<div class=\"output_area\">\n<div class=\"prompt\"><\/div>\n<div class=\"output_png output_subarea\"><img decoding=\"async\" alt=\"No description has been provided for this image\" src=\"data:image\/png;base64,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\" \/><\/div>\n<\/div>\n<div class=\"output_area\">\n<div class=\"prompt\"><\/div>\n<div class=\"output_png output_subarea\"><img decoding=\"async\" alt=\"No description has been provided for this image\" src=\"data:image\/png;base64,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\" \/><\/div>\n<\/div>\n<div class=\"output_area\">\n<div class=\"prompt\"><\/div>\n<div class=\"output_png output_subarea\"><img decoding=\"async\" alt=\"No description has been provided for this image\" src=\"data:image\/png;base64,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\" \/><\/div>\n<\/div>\n<div class=\"output_area\">\n<div class=\"prompt\"><\/div>\n<div class=\"output_png output_subarea\"><img decoding=\"async\" alt=\"No description has been provided for this image\" src=\"data:image\/png;base64,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\" \/><\/div>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing text_cell rendered\">\n<div class=\"inner_cell\">\n<div class=\"text_cell_render border-box-sizing rendered_html\">\n<p>\u4e0a\u306b\u793a\u3057\u305f\u5de6\u5074\u306e\u753b\u50cf\u306f\u3001\u884c\u5217 $W_1$ \u304c\u4fdd\u6301\u3057\u3066\u3044\u308b\u7279\u5fb4\u3092\u53ef\u8996\u5316\u3057\u305f\u3082\u306e\u3067\u3059\u3002\u3053\u308c\u306f\u3001\u5b66\u7fd2\u30c7\u30fc\u30bf\u306b\u73fe\u308c\u308b\u5171\u901a\u3057\u305f\u7279\u5fb4\u3092\u62bd\u51fa\u3057\u305f \u201c\u57fa\u5e95\u753b\u50cf\u201d \u306b\u76f8\u5f53\u3057\u3001\u30e2\u30c7\u30eb\u306f\u3053\u308c\u3089\u3092\u7d44\u307f\u5408\u308f\u305b\u308b\u3053\u3068\u3067\u5143\u306e\u753b\u50cf\u3092\u518d\u69cb\u6210\u3067\u304d\u308b\u3088\u3046\u306b\u5b66\u7fd2\u3057\u3066\u3044\u307e\u3059\u3002<\/p>\n<p>\u4e2d\u592e\u306e\u68d2\u30b0\u30e9\u30d5\u306f\u3001\u6a2a\u8ef8\u304c\u30e9\u30d9\u30eb\uff080\u301c9\uff09\u3001\u7e26\u8ef8\u304c\u4f7f\u7528\u56de\u6570\u3092\u8868\u3057\u3066\u304a\u308a\u3001\u3042\u308b\u57fa\u5e95\u753b\u50cf\u304c\u3069\u306e\u6570\u5b57\u3092\u518d\u69cb\u6210\u3059\u308b\u969b\u306b\u3069\u306e\u7a0b\u5ea6\u5229\u7528\u3055\u308c\u305f\u304b\u3092\u793a\u3057\u3066\u3044\u307e\u3059\u3002\u4f8b\u3048\u3070\u3001\u30e9\u30d9\u30eb0\u306e\u68d2\u304c20\u3067\u3042\u308c\u3070\u3001\u305d\u306e\u57fa\u5e95\u753b\u50cf\u306f\u300c0\u300d\u3068\u3044\u3046\u6570\u5b57\u3092\u518d\u69cb\u6210\u3059\u308b\u969b\u306b 20 \u56de\u5229\u7528\u3055\u308c\u305f\u3053\u3068\u3092\u610f\u5473\u3057\u307e\u3059\u3002<\/p>\n<p>\u53f3\u5074\u306e\u68d2\u30b0\u30e9\u30d5\u306f\u3001\u884c\u5217 $W_2$ \u306b\u542b\u307e\u308c\u308b\u30af\u30e9\u30b9\u60c5\u5831\u306e\u91cd\u307f\u3092\u53ef\u8996\u5316\u3057\u305f\u3082\u306e\u3067\u3059\u3002<br \/>\n$W_2$ \u306b\u306f\u3001\u5bfe\u5fdc\u3059\u308b\u57fa\u5e95\u753b\u50cf\u304c\u300c\u3069\u306e\u6570\u5b57\u3092\u518d\u69cb\u6210\u3057\u3084\u3059\u3044\uff08\uff1d\u3069\u306e\u30af\u30e9\u30b9\u3089\u3057\u3044\u7279\u5fb4\u3092\u6301\u3064\uff09\u300d\u304b\u3092\u53cd\u6620\u3057\u305f\u5024\u304c\u5b66\u7fd2\u3055\u308c\u3066\u304a\u308a\u3001\u57fa\u5e95\u753b\u50cf\u306e\u6301\u3064\u30af\u30e9\u30b9\u50be\u5411\u3092\u793a\u3059\u6307\u6a19\u3068\u306a\u308a\u307e\u3059\u3002<\/p>\n<p>\u4e2d\u592e\u306e\u4f7f\u7528\u56de\u6570\u306e\u5206\u5e03\u3068\u3001\u53f3\u5074\u306e\u30af\u30e9\u30b9\u91cd\u307f\u304c\u6982\u306d\u4e00\u81f4\u3057\u3066\u3044\u308b\u5834\u5408\u3001\u305d\u306e\u57fa\u5e95\u753b\u50cf\u304c\u5b9f\u969b\u306b\u518d\u69cb\u6210\u306b\u4f7f\u308f\u308c\u305f\u6570\u5b57\u3068\u3001\u30e2\u30c7\u30eb\u304c\u305d\u306e\u57fa\u5e95\u753b\u50cf\u3092\u300c\u3069\u306e\u30af\u30e9\u30b9\u3089\u3057\u3044\u300d\u3068\u5224\u65ad\u3057\u3066\u3044\u308b\u60c5\u5831\u304c\u4e00\u81f4\u3057\u3066\u3044\u308b\u3053\u3068\u306b\u306a\u308a\u3001\u5b66\u7fd2\u304c\u3046\u307e\u304f\u9032\u3093\u3067\u3044\u308b\u3068\u5224\u65ad\u3067\u304d\u307e\u3059\u3002<br \/>\n\u57fa\u5e95\u753b\u50cf\u306e\u4e2d\u306b\u306f\u76f4\u611f\u7684\u306b\u5206\u304b\u308a\u306b\u304f\u3044\u3082\u306e\u3082\u3042\u308a\u307e\u3059\u304c\u3001\u4f8b\u3048\u3070\u3001\u5de6\u306e\u753b\u50cf\u304c\u300c3\u300d\u306b\u898b\u3048\u308b\u5f62\u72b6\u3092\u3057\u3066\u3044\u308b\u5834\u5408\u3001\u68d2\u30b0\u30e9\u30d5\u3067\u3082\u30e9\u30d9\u30eb3\u306e\u5024\u304c\u9ad8\u304f\u306a\u308b\u306a\u3069\u3001\u8996\u899a\u7684\u306a\u5370\u8c61\u3068\u30e2\u30c7\u30eb\u306e\u5b66\u7fd2\u7d50\u679c\u304c\u6574\u5408\u3059\u308b\u4f8b\u3082\u78ba\u8a8d\u3067\u304d\u307e\u3057\u305f\u3002<\/p>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing text_cell rendered\">\n<div class=\"inner_cell\">\n<div class=\"text_cell_render border-box-sizing rendered_html\">\n<h3><span class=\"ez-toc-section\" id=\"W%E3%82%92%E7%94%A8%E3%81%84%E3%81%9F%E5%86%8D%E6%A7%8B%E6%88%90\"><\/span>$W$\u3092\u7528\u3044\u305f\u518d\u69cb\u6210<span class=\"ez-toc-section-end\"><\/span><\/h3>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing text_cell rendered\">\n<div class=\"inner_cell\">\n<div class=\"text_cell_render border-box-sizing rendered_html\">\n<p>\u30a8\u30dd\u30c3\u30af\u6570\u306b\u5bfe\u3059\u308b\u7cbe\u5ea6\u3068\u4ea4\u5dee\u30a8\u30f3\u30c8\u30ed\u30d4\u30fc\u8aa4\u5dee\u306e\u6bd4\u8f03\u5b9f\u9a13\u306e\u7d50\u679c\u3092\u7528\u3044\u3066\u3001\u5b66\u7fd2\u30c7\u30fc\u30bf\u3092\u518d\u69cb\u6210\u3057\u3001\u5b66\u7fd2\u304c\u3069\u306e\u7a0b\u5ea6\u3046\u307e\u304f\u3044\u3063\u3066\u3044\u308b\u304b\u3092$W$\u306b\u3088\u308b\u518d\u69cb\u6210\u3068\u3044\u3046\u89b3\u70b9\u304b\u3089\u8a55\u4fa1\u3057\u307e\u3059\u3002\u4ee5\u4e0b\u3067\u306f\u307e\u305a\u3001$W$\u3092\u7528\u3044\u3066\u5b66\u7fd2\u30c7\u30fc\u30bf\u3092\u518d\u69cb\u6210\u3057\u3001\u305d\u306e\u753b\u50cf\u3068\u5143\u306e\u5b66\u7fd2\u30c7\u30fc\u30bf\u306e\u753b\u50cf\u3092\u4e26\u3079\u3066\u8868\u793a\u3059\u308b\u95a2\u6570\u3092\u4f5c\u6210\u3057\u307e\u3059\u3002<\/p>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing code_cell rendered\">\n<div class=\"input\">\n<div class=\"inner_cell\">\n<div class=\"input_area\">\n<div class=\"highlight hl-python\">\n<pre><span><\/span><span class=\"k\">def<\/span><span class=\"w\"> <\/span><span class=\"nf\">reconstruct_image_auto<\/span><span class=\"p\">(<\/span><span class=\"n\">W<\/span><span class=\"p\">,<\/span> <span class=\"n\">H_test<\/span><span class=\"p\">,<\/span> <span class=\"n\">visualize_idx<\/span><span class=\"p\">):<\/span>\r\n<span class=\"w\">    <\/span><span class=\"sd\">\"\"\"<\/span>\r\n<span class=\"sd\">    \u30c6\u30b9\u30c8\u753b\u50cf\u3092\u3001H_test\u306e\u975e\u30bc\u30ed\u6210\u5206\u306e\u307f\u3092\u7528\u3044\u3066\u518d\u69cb\u6210\u3059\u308b<\/span>\r\n<span class=\"sd\">    &lt;\u5165\u529b&gt;<\/span>\r\n<span class=\"sd\">    W\uff1a(n_features, k), \u57fa\u5e95\u884c\u5217<\/span>\r\n<span class=\"sd\">    H_test\uff1a(k, n_test), \u30c6\u30b9\u30c8\u30c7\u30fc\u30bf\u306e\u4fc2\u6570\u884c\u5217<\/span>\r\n<span class=\"sd\">    visualize_idx\uff1a(int), \u5bfe\u8c61\u306e\u30c6\u30b9\u30c8\u30b5\u30f3\u30d7\u30eb\u30a4\u30f3\u30c7\u30c3\u30af\u30b9<\/span>\r\n<span class=\"sd\">    &lt;\u51fa\u529b&gt;<\/span>\r\n<span class=\"sd\">    recon\uff1a(784,), \u518d\u69cb\u6210\u3055\u308c\u305f\u753b\u50cf\u30d9\u30af\u30c8\u30eb\uff08\u6700\u521d\u306e784\u6210\u5206\u306e\u307f\uff09<\/span>\r\n<span class=\"sd\">    active\uff1a(array), \u518d\u69cb\u6210\u306b\u4f7f\u7528\u3055\u308c\u305f\u57fa\u5e95\u306e\u30a4\u30f3\u30c7\u30c3\u30af\u30b9\u914d\u5217<\/span>\r\n<span class=\"sd\">    \"\"\"<\/span>\r\n\r\n    <span class=\"n\">active<\/span> <span class=\"o\">=<\/span> <span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">where<\/span><span class=\"p\">(<\/span><span class=\"n\">H_test<\/span><span class=\"p\">[:,<\/span> <span class=\"n\">visualize_idx<\/span><span class=\"p\">]<\/span> <span class=\"o\">&gt;<\/span> <span class=\"mi\">0<\/span><span class=\"p\">)[<\/span><span class=\"mi\">0<\/span><span class=\"p\">]<\/span>\r\n\r\n    <span class=\"n\">h<\/span> <span class=\"o\">=<\/span> <span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">zeros_like<\/span><span class=\"p\">(<\/span><span class=\"n\">H_test<\/span><span class=\"p\">[:,<\/span> <span class=\"n\">visualize_idx<\/span><span class=\"p\">])<\/span>\r\n    <span class=\"n\">h<\/span><span class=\"p\">[<\/span><span class=\"n\">active<\/span><span class=\"p\">]<\/span> <span class=\"o\">=<\/span> <span class=\"n\">H_test<\/span><span class=\"p\">[<\/span><span class=\"n\">active<\/span><span class=\"p\">,<\/span> <span class=\"n\">visualize_idx<\/span><span class=\"p\">]<\/span>\r\n\r\n    <span class=\"n\">recon<\/span> <span class=\"o\">=<\/span> <span class=\"n\">W<\/span> <span class=\"o\">@<\/span> <span class=\"n\">h<\/span>\r\n\r\n    <span class=\"k\">return<\/span> <span class=\"n\">recon<\/span><span class=\"p\">[:<\/span><span class=\"mi\">784<\/span><span class=\"p\">],<\/span> <span class=\"n\">active<\/span>\r\n\r\n<span class=\"k\">def<\/span><span class=\"w\"> <\/span><span class=\"nf\">show_reconstruction_auto<\/span><span class=\"p\">(<\/span><span class=\"n\">W<\/span><span class=\"p\">,<\/span> <span class=\"n\">H_test<\/span><span class=\"p\">,<\/span> <span class=\"n\">X_test<\/span><span class=\"p\">,<\/span> <span class=\"n\">visualize_idx<\/span><span class=\"p\">,<\/span> <span class=\"n\">y_test<\/span><span class=\"p\">,<\/span> <span class=\"n\">y_pred<\/span><span class=\"p\">):<\/span>\r\n<span class=\"w\">    <\/span><span class=\"sd\">\"\"\"<\/span>\r\n<span class=\"sd\">    \u5143\u306e\u753b\u50cf\u3068\u3001\u9078\u629e\u3055\u308c\u305f\u57fa\u5e95\u306b\u3088\u308a\u518d\u69cb\u6210\u3055\u308c\u305f\u753b\u50cf\u3092\u4e26\u3079\u3066\u8868\u793a\u3059\u308b<\/span>\r\n<span class=\"sd\">    &lt;\u5165\u529b&gt;<\/span>\r\n<span class=\"sd\">    W\uff1a(n_features, k), \u57fa\u5e95\u884c\u5217<\/span>\r\n<span class=\"sd\">    H_test\uff1a(k, n_test), \u30c6\u30b9\u30c8\u30c7\u30fc\u30bf\u306e\u4fc2\u6570\u884c\u5217<\/span>\r\n<span class=\"sd\">    X_test\uff1a(n_test, 784), \u30c6\u30b9\u30c8\u753b\u50cf\u30c7\u30fc\u30bf<\/span>\r\n<span class=\"sd\">    visualize_idx\uff1a(int), \u8868\u793a\u5bfe\u8c61\u306e\u30c6\u30b9\u30c8\u30b5\u30f3\u30d7\u30eb\u30a4\u30f3\u30c7\u30c3\u30af\u30b9<\/span>\r\n<span class=\"sd\">    &lt;\u51fa\u529b&gt;<\/span>\r\n<span class=\"sd\">    \u306a\u3057<\/span>\r\n<span class=\"sd\">    \"\"\"<\/span>\r\n    <span class=\"n\">recon<\/span><span class=\"p\">,<\/span> <span class=\"n\">active<\/span> <span class=\"o\">=<\/span> <span class=\"n\">reconstruct_image_auto<\/span><span class=\"p\">(<\/span><span class=\"n\">W<\/span><span class=\"p\">,<\/span> <span class=\"n\">H_test<\/span><span class=\"p\">,<\/span> <span class=\"n\">visualize_idx<\/span><span class=\"p\">)<\/span>\r\n\r\n    <span class=\"n\">fig<\/span><span class=\"p\">,<\/span> <span class=\"n\">axes<\/span> <span class=\"o\">=<\/span> <span class=\"n\">plt<\/span><span class=\"o\">.<\/span><span class=\"n\">subplots<\/span><span class=\"p\">(<\/span><span class=\"mi\">1<\/span><span class=\"p\">,<\/span> <span class=\"mi\">2<\/span><span class=\"p\">,<\/span> <span class=\"n\">figsize<\/span><span class=\"o\">=<\/span><span class=\"p\">(<\/span><span class=\"mi\">6<\/span><span class=\"p\">,<\/span> <span class=\"mi\">3<\/span><span class=\"p\">))<\/span>\r\n\r\n    <span class=\"n\">axes<\/span><span class=\"p\">[<\/span><span class=\"mi\">0<\/span><span class=\"p\">]<\/span><span class=\"o\">.<\/span><span class=\"n\">imshow<\/span><span class=\"p\">(<\/span><span class=\"n\">X_test<\/span><span class=\"p\">[<\/span><span class=\"n\">visualize_idx<\/span><span class=\"p\">]<\/span><span class=\"o\">.<\/span><span class=\"n\">reshape<\/span><span class=\"p\">(<\/span><span class=\"mi\">28<\/span><span class=\"p\">,<\/span> <span class=\"mi\">28<\/span><span class=\"p\">),<\/span> <span class=\"n\">cmap<\/span><span class=\"o\">=<\/span><span class=\"s1\">'gray'<\/span><span class=\"p\">)<\/span>\r\n    <span class=\"n\">axes<\/span><span class=\"p\">[<\/span><span class=\"mi\">0<\/span><span class=\"p\">]<\/span><span class=\"o\">.<\/span><span class=\"n\">set_title<\/span><span class=\"p\">(<\/span><span class=\"sa\">f<\/span><span class=\"s2\">\"Original<\/span><span class=\"se\">\\n<\/span><span class=\"s2\">(True: <\/span><span class=\"si\">{<\/span><span class=\"n\">y_test<\/span><span class=\"p\">[<\/span><span class=\"n\">visualize_idx<\/span><span class=\"p\">]<\/span><span class=\"si\">}<\/span><span class=\"s2\">)\"<\/span><span class=\"p\">)<\/span>\r\n    <span class=\"n\">axes<\/span><span class=\"p\">[<\/span><span class=\"mi\">0<\/span><span class=\"p\">]<\/span><span class=\"o\">.<\/span><span class=\"n\">axis<\/span><span class=\"p\">(<\/span><span class=\"s2\">\"off\"<\/span><span class=\"p\">)<\/span>\r\n\r\n    <span class=\"n\">axes<\/span><span class=\"p\">[<\/span><span class=\"mi\">1<\/span><span class=\"p\">]<\/span><span class=\"o\">.<\/span><span class=\"n\">imshow<\/span><span class=\"p\">(<\/span><span class=\"n\">recon<\/span><span class=\"o\">.<\/span><span class=\"n\">reshape<\/span><span class=\"p\">(<\/span><span class=\"mi\">28<\/span><span class=\"p\">,<\/span> <span class=\"mi\">28<\/span><span class=\"p\">),<\/span> <span class=\"n\">cmap<\/span><span class=\"o\">=<\/span><span class=\"s1\">'gray'<\/span><span class=\"p\">)<\/span>\r\n    <span class=\"n\">axes<\/span><span class=\"p\">[<\/span><span class=\"mi\">1<\/span><span class=\"p\">]<\/span><span class=\"o\">.<\/span><span class=\"n\">set_title<\/span><span class=\"p\">(<\/span><span class=\"sa\">f<\/span><span class=\"s2\">\"Reconstructed<\/span><span class=\"se\">\\n<\/span><span class=\"s2\">(bases: <\/span><span class=\"si\">{<\/span><span class=\"nb\">len<\/span><span class=\"p\">(<\/span><span class=\"n\">active<\/span><span class=\"p\">)<\/span><span class=\"si\">}<\/span><span class=\"s2\">)<\/span><span class=\"se\">\\n<\/span><span class=\"s2\">(Pred: <\/span><span class=\"si\">{<\/span><span class=\"n\">y_pred<\/span><span class=\"p\">[<\/span><span class=\"n\">visualize_idx<\/span><span class=\"p\">]<\/span><span class=\"si\">}<\/span><span class=\"s2\">)\"<\/span><span class=\"p\">)<\/span>\r\n    <span class=\"n\">axes<\/span><span class=\"p\">[<\/span><span class=\"mi\">1<\/span><span class=\"p\">]<\/span><span class=\"o\">.<\/span><span class=\"n\">axis<\/span><span class=\"p\">(<\/span><span class=\"s2\">\"off\"<\/span><span class=\"p\">)<\/span>\r\n\r\n    <span class=\"n\">plt<\/span><span class=\"o\">.<\/span><span class=\"n\">tight_layout<\/span><span class=\"p\">()<\/span>\r\n    <span class=\"n\">plt<\/span><span class=\"o\">.<\/span><span class=\"n\">show<\/span><span class=\"p\">()<\/span>\r\n\r\n    <span class=\"nb\">print<\/span><span class=\"p\">(<\/span><span class=\"s2\">\"Used basis columns:\"<\/span><span class=\"p\">,<\/span> <span class=\"n\">active<\/span><span class=\"p\">)<\/span>\r\n<\/pre>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing text_cell rendered\">\n<div class=\"inner_cell\">\n<div class=\"text_cell_render border-box-sizing rendered_html\">\n<p>\u4f5c\u6210\u3057\u305f\u95a2\u6570\u3092\u7528\u3044\u3066\u518d\u69cb\u6210\u753b\u50cf\u3092\u8868\u793a\u3057\u307e\u3059\u3002<\/p>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing code_cell rendered\">\n<div class=\"input\">\n<div class=\"inner_cell\">\n<div class=\"input_area\">\n<div class=\"highlight hl-python\">\n<pre><span><\/span><span class=\"nb\">print<\/span><span class=\"p\">(<\/span><span class=\"s2\">\"=== \u518d\u69cb\u6210\u753b\u50cf\u306e\u8868\u793a ===\"<\/span><span class=\"p\">)<\/span>\r\n\r\n<span class=\"n\">visualize_idx<\/span> <span class=\"o\">=<\/span> <span class=\"mi\">10<\/span>  <span class=\"c1\"># \u8868\u793a\u3055\u305b\u308b\u30a4\u30f3\u30c7\u30c3\u30af\u30b9<\/span>\r\n<span class=\"n\">show_reconstruction_auto<\/span><span class=\"p\">(<\/span><span class=\"n\">W_final<\/span><span class=\"p\">,<\/span> <span class=\"n\">H_test<\/span><span class=\"p\">,<\/span> <span class=\"n\">X_test<\/span><span class=\"p\">,<\/span> <span class=\"nb\">min<\/span><span class=\"p\">(<\/span><span class=\"n\">m_test<\/span><span class=\"o\">-<\/span><span class=\"mi\">1<\/span><span class=\"p\">,<\/span> <span class=\"n\">visualize_idx<\/span><span class=\"p\">),<\/span> <span class=\"n\">y_test<\/span><span class=\"p\">,<\/span> <span class=\"n\">y_pred<\/span><span class=\"p\">)<\/span>\r\n<\/pre>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"output_wrapper\">\n<div class=\"output\">\n<div class=\"output_area\">\n<div class=\"prompt\"><\/div>\n<div class=\"output_subarea output_stream output_stdout output_text\">\n<pre>=== \u518d\u69cb\u6210\u753b\u50cf\u306e\u8868\u793a ===\r\n<\/pre>\n<\/div>\n<\/div>\n<div class=\"output_area\">\n<div class=\"prompt\"><\/div>\n<div class=\"output_png output_subarea\"><img decoding=\"async\" alt=\"No description has been provided for this image\" src=\"data:image\/png;base64,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\" \/><\/div>\n<\/div>\n<div class=\"output_area\">\n<div class=\"prompt\"><\/div>\n<div class=\"output_subarea output_stream output_stdout output_text\">\n<pre>Used basis columns: [ 2  7  8  9 15 18 20 23 24 26 27 30 31 34 36 38 39]\r\n<\/pre>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing text_cell rendered\">\n<div class=\"inner_cell\">\n<div class=\"text_cell_render border-box-sizing rendered_html\">\n<p>\u7570\u306a\u308bvisualize_idx\u3067\u69d8\u3005\u306a\u753b\u50cf\u3092\u8868\u793a\u3055\u305b\u308b\u3068\u3001\u5143\u753b\u50cf\u3092\u7cbe\u5ea6\u3088\u304f\u518d\u73fe\u3057\u3066\u3044\u308b\u518d\u69cb\u6210\u753b\u50cf\u304c\u8907\u6570\u78ba\u8a8d\u3067\u304d\u307e\u3059\u3002\u3053\u306e\u3053\u3068\u304b\u3089\u3001\u518d\u69cb\u6210\u3068\u3044\u3046\u89b3\u70b9\u304b\u3089\u3082NBMF\u306b\u3088\u308b\u5b66\u7fd2\u306f\u3046\u307e\u304f\u3044\u3063\u3066\u3044\u308b\u3068\u8a00\u3048\u307e\u3059\u3002<\/p>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing text_cell rendered\">\n<div class=\"inner_cell\">\n<div class=\"text_cell_render border-box-sizing rendered_html\">\n<h3><span class=\"ez-toc-section\" id=\"%E8%AA%A4%E5%88%86%E9%A1%9E%E3%81%AB%E5%AF%BE%E3%81%99%E3%82%8B%E5%88%86%E6%9E%90\"><\/span>\u8aa4\u5206\u985e\u306b\u5bfe\u3059\u308b\u5206\u6790<span class=\"ez-toc-section-end\"><\/span><\/h3>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing text_cell rendered\">\n<div class=\"inner_cell\">\n<div class=\"text_cell_render border-box-sizing rendered_html\">\n<p>\u30a8\u30dd\u30c3\u30af\u6570\u306b\u5bfe\u3059\u308b\u7cbe\u5ea6\u3068\u4ea4\u5dee\u30a8\u30f3\u30c8\u30ed\u30d4\u30fc\u8aa4\u5dee\u306e\u6bd4\u8f03\u5b9f\u9a13\u306e\u7d50\u679c\u306b\u304a\u3044\u3066\u3001\u8aa4\u5206\u985e\u304c\u751f\u3058\u3066\u3057\u307e\u3063\u305f\u30b5\u30f3\u30d7\u30eb\u306e\u5206\u6790\u3092\u884c\u3044\u307e\u3059\u3002\u4ee5\u4e0b\u3067\u306f\u3001\u8aa4\u5206\u985e\u304c\u751f\u3058\u305f\u30b5\u30f3\u30d7\u30eb\u306e\u753b\u50cf\u3068\u4e88\u6e2c\u3057\u305f\u30e9\u30d9\u30eb\u3092\u8868\u793a\u3059\u308b\u95a2\u6570\u3068\u3001\u3069\u306e\u30e9\u30d9\u30eb\u3068\u3069\u306e\u30e9\u30d9\u30eb\u3067\u8aa4\u5206\u985e\u304c\u751f\u3058\u305f\u304b\u3092\u8868\u3059\u6df7\u540c\u884c\u5217\u3092\u30d2\u30fc\u30c8\u30de\u30c3\u30d7\u3068\u3057\u3066\u8868\u793a\u3059\u308b\u95a2\u6570\u3092\u4f5c\u6210\u3057\u307e\u3059\u3002<\/p>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing code_cell rendered\">\n<div class=\"input\">\n<div class=\"inner_cell\">\n<div class=\"input_area\">\n<div class=\"highlight hl-python\">\n<pre><span><\/span><span class=\"k\">def<\/span><span class=\"w\"> <\/span><span class=\"nf\">show_misclassified<\/span><span class=\"p\">(<\/span><span class=\"n\">X_test<\/span><span class=\"p\">,<\/span> <span class=\"n\">y_test<\/span><span class=\"p\">,<\/span> <span class=\"n\">y_pred<\/span><span class=\"p\">,<\/span> <span class=\"n\">num<\/span><span class=\"o\">=<\/span><span class=\"mi\">10<\/span><span class=\"p\">):<\/span>\r\n<span class=\"w\">    <\/span><span class=\"sd\">\"\"\"<\/span>\r\n<span class=\"sd\">    \u5206\u985e\u3092\u9593\u9055\u3048\u305f\u30c6\u30b9\u30c8\u753b\u50cf\u3092\u3001\u6307\u5b9a\u3055\u308c\u305f\u679a\u6570\u5206\u3060\u3051\u4e26\u3079\u3066\u8868\u793a\u3059\u308b<\/span>\r\n<span class=\"sd\">    &lt;\u5165\u529b&gt;<\/span>\r\n<span class=\"sd\">    X_test\uff1a(n_test, 784), \u30c6\u30b9\u30c8\u753b\u50cf\u30c7\u30fc\u30bf<\/span>\r\n<span class=\"sd\">    y_test\uff1a(n_test,), \u6b63\u89e3\u30e9\u30d9\u30eb<\/span>\r\n<span class=\"sd\">    y_pred\uff1a(n_test,), \u4e88\u6e2c\u30e9\u30d9\u30eb<\/span>\r\n<span class=\"sd\">    num\uff1a(int), \u8868\u793a\u3059\u308b\u753b\u50cf\u306e\u679a\u6570\uff08\u30c7\u30d5\u30a9\u30eb\u30c810\uff09<\/span>\r\n<span class=\"sd\">    &lt;\u51fa\u529b&gt;<\/span>\r\n<span class=\"sd\">    \u306a\u3057<\/span>\r\n<span class=\"sd\">    \"\"\"<\/span>\r\n    <span class=\"n\">wrong<\/span> <span class=\"o\">=<\/span> <span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">where<\/span><span class=\"p\">(<\/span><span class=\"n\">y_test<\/span> <span class=\"o\">!=<\/span> <span class=\"n\">y_pred<\/span><span class=\"p\">)[<\/span><span class=\"mi\">0<\/span><span class=\"p\">]<\/span>\r\n    <span class=\"n\">show_idx<\/span> <span class=\"o\">=<\/span> <span class=\"n\">wrong<\/span><span class=\"p\">[:<\/span><span class=\"n\">num<\/span><span class=\"p\">]<\/span>\r\n\r\n    <span class=\"n\">fig<\/span><span class=\"p\">,<\/span> <span class=\"n\">axes<\/span> <span class=\"o\">=<\/span> <span class=\"n\">plt<\/span><span class=\"o\">.<\/span><span class=\"n\">subplots<\/span><span class=\"p\">(<\/span><span class=\"mi\">1<\/span><span class=\"p\">,<\/span> <span class=\"nb\">len<\/span><span class=\"p\">(<\/span><span class=\"n\">show_idx<\/span><span class=\"p\">),<\/span> <span class=\"n\">figsize<\/span><span class=\"o\">=<\/span><span class=\"p\">(<\/span><span class=\"mf\">1.8<\/span><span class=\"o\">*<\/span><span class=\"nb\">len<\/span><span class=\"p\">(<\/span><span class=\"n\">show_idx<\/span><span class=\"p\">),<\/span> <span class=\"mi\">2<\/span><span class=\"p\">))<\/span>\r\n\r\n    <span class=\"k\">if<\/span> <span class=\"nb\">len<\/span><span class=\"p\">(<\/span><span class=\"n\">show_idx<\/span><span class=\"p\">)<\/span> <span class=\"o\">==<\/span> <span class=\"mi\">1<\/span><span class=\"p\">:<\/span>\r\n        <span class=\"n\">axes<\/span> <span class=\"o\">=<\/span> <span class=\"p\">[<\/span><span class=\"n\">axes<\/span><span class=\"p\">]<\/span>\r\n\r\n    <span class=\"k\">for<\/span> <span class=\"n\">ax<\/span><span class=\"p\">,<\/span> <span class=\"n\">i<\/span> <span class=\"ow\">in<\/span> <span class=\"nb\">zip<\/span><span class=\"p\">(<\/span><span class=\"n\">axes<\/span><span class=\"p\">,<\/span> <span class=\"n\">show_idx<\/span><span class=\"p\">):<\/span>\r\n        <span class=\"n\">ax<\/span><span class=\"o\">.<\/span><span class=\"n\">imshow<\/span><span class=\"p\">(<\/span><span class=\"n\">X_test<\/span><span class=\"p\">[<\/span><span class=\"n\">i<\/span><span class=\"p\">]<\/span><span class=\"o\">.<\/span><span class=\"n\">reshape<\/span><span class=\"p\">(<\/span><span class=\"mi\">28<\/span><span class=\"p\">,<\/span> <span class=\"mi\">28<\/span><span class=\"p\">),<\/span> <span class=\"n\">cmap<\/span><span class=\"o\">=<\/span><span class=\"s1\">'gray'<\/span><span class=\"p\">)<\/span>\r\n        <span class=\"n\">ax<\/span><span class=\"o\">.<\/span><span class=\"n\">set_title<\/span><span class=\"p\">(<\/span><span class=\"sa\">f<\/span><span class=\"s2\">\"T=<\/span><span class=\"si\">{<\/span><span class=\"n\">y_test<\/span><span class=\"p\">[<\/span><span class=\"n\">i<\/span><span class=\"p\">]<\/span><span class=\"si\">}<\/span><span class=\"s2\">, P=<\/span><span class=\"si\">{<\/span><span class=\"n\">y_pred<\/span><span class=\"p\">[<\/span><span class=\"n\">i<\/span><span class=\"p\">]<\/span><span class=\"si\">}<\/span><span class=\"s2\">\"<\/span><span class=\"p\">)<\/span>\r\n        <span class=\"n\">ax<\/span><span class=\"o\">.<\/span><span class=\"n\">axis<\/span><span class=\"p\">(<\/span><span class=\"s1\">'off'<\/span><span class=\"p\">)<\/span>\r\n\r\n    <span class=\"n\">plt<\/span><span class=\"o\">.<\/span><span class=\"n\">tight_layout<\/span><span class=\"p\">()<\/span>\r\n    <span class=\"n\">plt<\/span><span class=\"o\">.<\/span><span class=\"n\">show<\/span><span class=\"p\">()<\/span>\r\n\r\n<span class=\"k\">def<\/span><span class=\"w\"> <\/span><span class=\"nf\">plot_normalized_confusion_matrix_heatmap<\/span><span class=\"p\">(<\/span><span class=\"n\">y_True<\/span><span class=\"p\">,<\/span> <span class=\"n\">y_pred<\/span><span class=\"p\">):<\/span>\r\n<span class=\"w\">    <\/span><span class=\"sd\">\"\"\"<\/span>\r\n<span class=\"sd\">    \u6b63\u89e3\u30e9\u30d9\u30eb\u3068\u4e88\u6e2c\u30e9\u30d9\u30eb\u304b\u3089\u6df7\u540c\u884c\u5217\u3092\u4f5c\u6210\u3057\u3001<\/span>\r\n<span class=\"sd\">    \u6b63\u89e3\u30e9\u30d9\u30eb\u3054\u3068\u306e\u5272\u5408\uff08\u6b63\u898f\u5316\uff09\u306b\u3057\u3066\u30d2\u30fc\u30c8\u30de\u30c3\u30d7\u3068\u3057\u3066\u8868\u793a\u3059\u308b<\/span>\r\n<span class=\"sd\">    &lt;\u5165\u529b&gt;<\/span>\r\n<span class=\"sd\">    y_True\uff1a(n_samples,), \u6b63\u89e3\u30e9\u30d9\u30eb<\/span>\r\n<span class=\"sd\">    y_pred\uff1a(n_samples,), \u4e88\u6e2c\u30e9\u30d9\u30eb<\/span>\r\n<span class=\"sd\">    &lt;\u51fa\u529b&gt;<\/span>\r\n<span class=\"sd\">    cm_norm\uff1a(10, 10), \u6b63\u898f\u5316\u3055\u308c\u305f\u6df7\u540c\u884c\u5217<\/span>\r\n<span class=\"sd\">    \"\"\"<\/span>\r\n\r\n    <span class=\"n\">cm<\/span> <span class=\"o\">=<\/span> <span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">zeros<\/span><span class=\"p\">((<\/span><span class=\"mi\">10<\/span><span class=\"p\">,<\/span> <span class=\"mi\">10<\/span><span class=\"p\">),<\/span> <span class=\"n\">dtype<\/span><span class=\"o\">=<\/span><span class=\"nb\">int<\/span><span class=\"p\">)<\/span>\r\n    <span class=\"k\">for<\/span> <span class=\"n\">t<\/span><span class=\"p\">,<\/span> <span class=\"n\">p<\/span> <span class=\"ow\">in<\/span> <span class=\"nb\">zip<\/span><span class=\"p\">(<\/span><span class=\"n\">y_True<\/span><span class=\"p\">,<\/span> <span class=\"n\">y_pred<\/span><span class=\"p\">):<\/span>\r\n        <span class=\"n\">cm<\/span><span class=\"p\">[<\/span><span class=\"n\">t<\/span><span class=\"p\">,<\/span> <span class=\"n\">p<\/span><span class=\"p\">]<\/span> <span class=\"o\">+=<\/span> <span class=\"mi\">1<\/span>\r\n\r\n    <span class=\"n\">row_sums<\/span> <span class=\"o\">=<\/span> <span class=\"n\">cm<\/span><span class=\"o\">.<\/span><span class=\"n\">sum<\/span><span class=\"p\">(<\/span><span class=\"n\">axis<\/span><span class=\"o\">=<\/span><span class=\"mi\">1<\/span><span class=\"p\">,<\/span> <span class=\"n\">keepdims<\/span><span class=\"o\">=<\/span><span class=\"kc\">True<\/span><span class=\"p\">)<\/span>\r\n\r\n    <span class=\"n\">row_sums<\/span><span class=\"p\">[<\/span><span class=\"n\">row_sums<\/span> <span class=\"o\">==<\/span> <span class=\"mi\">0<\/span><span class=\"p\">]<\/span> <span class=\"o\">=<\/span> <span class=\"mi\">1<\/span>\r\n\r\n    <span class=\"n\">cm_norm<\/span> <span class=\"o\">=<\/span> <span class=\"n\">cm<\/span> <span class=\"o\">\/<\/span> <span class=\"n\">row_sums<\/span>\r\n\r\n    <span class=\"n\">plt<\/span><span class=\"o\">.<\/span><span class=\"n\">figure<\/span><span class=\"p\">(<\/span><span class=\"n\">figsize<\/span><span class=\"o\">=<\/span><span class=\"p\">(<\/span><span class=\"mi\">8<\/span><span class=\"p\">,<\/span> <span class=\"mi\">7<\/span><span class=\"p\">))<\/span>\r\n    <span class=\"n\">plt<\/span><span class=\"o\">.<\/span><span class=\"n\">imshow<\/span><span class=\"p\">(<\/span><span class=\"n\">cm_norm<\/span><span class=\"p\">,<\/span> <span class=\"n\">interpolation<\/span><span class=\"o\">=<\/span><span class=\"s2\">\"nearest\"<\/span><span class=\"p\">,<\/span> <span class=\"n\">vmin<\/span><span class=\"o\">=<\/span><span class=\"mi\">0<\/span><span class=\"p\">,<\/span> <span class=\"n\">vmax<\/span><span class=\"o\">=<\/span><span class=\"mi\">1<\/span><span class=\"p\">)<\/span>\r\n    <span class=\"n\">plt<\/span><span class=\"o\">.<\/span><span class=\"n\">title<\/span><span class=\"p\">(<\/span><span class=\"s2\">\"Normalized Confusion Matrix\"<\/span><span class=\"p\">)<\/span>\r\n    <span class=\"n\">plt<\/span><span class=\"o\">.<\/span><span class=\"n\">xlabel<\/span><span class=\"p\">(<\/span><span class=\"s2\">\"Predicted Label\"<\/span><span class=\"p\">)<\/span>\r\n    <span class=\"n\">plt<\/span><span class=\"o\">.<\/span><span class=\"n\">ylabel<\/span><span class=\"p\">(<\/span><span class=\"s2\">\"True Label\"<\/span><span class=\"p\">)<\/span>\r\n\r\n    <span class=\"n\">plt<\/span><span class=\"o\">.<\/span><span class=\"n\">colorbar<\/span><span class=\"p\">()<\/span>\r\n\r\n    <span class=\"k\">for<\/span> <span class=\"n\">i<\/span> <span class=\"ow\">in<\/span> <span class=\"nb\">range<\/span><span class=\"p\">(<\/span><span class=\"mi\">10<\/span><span class=\"p\">):<\/span>\r\n        <span class=\"k\">for<\/span> <span class=\"n\">j<\/span> <span class=\"ow\">in<\/span> <span class=\"nb\">range<\/span><span class=\"p\">(<\/span><span class=\"mi\">10<\/span><span class=\"p\">):<\/span>\r\n            <span class=\"n\">text_val<\/span> <span class=\"o\">=<\/span> <span class=\"sa\">f<\/span><span class=\"s2\">\"<\/span><span class=\"si\">{<\/span><span class=\"n\">cm_norm<\/span><span class=\"p\">[<\/span><span class=\"n\">i<\/span><span class=\"p\">,<\/span><span class=\"w\"> <\/span><span class=\"n\">j<\/span><span class=\"p\">]<\/span><span class=\"si\">:<\/span><span class=\"s2\">.2f<\/span><span class=\"si\">}<\/span><span class=\"s2\">\"<\/span>\r\n\r\n            <span class=\"n\">plt<\/span><span class=\"o\">.<\/span><span class=\"n\">text<\/span><span class=\"p\">(<\/span><span class=\"n\">j<\/span><span class=\"p\">,<\/span> <span class=\"n\">i<\/span><span class=\"p\">,<\/span> <span class=\"n\">text_val<\/span><span class=\"p\">,<\/span>\r\n                     <span class=\"n\">ha<\/span><span class=\"o\">=<\/span><span class=\"s1\">'center'<\/span><span class=\"p\">,<\/span> <span class=\"n\">va<\/span><span class=\"o\">=<\/span><span class=\"s1\">'center'<\/span><span class=\"p\">,<\/span>\r\n                     <span class=\"n\">color<\/span><span class=\"o\">=<\/span><span class=\"s1\">'white'<\/span> <span class=\"k\">if<\/span> <span class=\"n\">cm_norm<\/span><span class=\"p\">[<\/span><span class=\"n\">i<\/span><span class=\"p\">,<\/span> <span class=\"n\">j<\/span><span class=\"p\">]<\/span> <span class=\"o\">&gt;<\/span> <span class=\"mf\">0.5<\/span> <span class=\"k\">else<\/span> <span class=\"s1\">'black'<\/span><span class=\"p\">,<\/span>\r\n                     <span class=\"n\">fontsize<\/span><span class=\"o\">=<\/span><span class=\"mi\">9<\/span><span class=\"p\">)<\/span>\r\n\r\n    <span class=\"n\">plt<\/span><span class=\"o\">.<\/span><span class=\"n\">xticks<\/span><span class=\"p\">(<\/span><span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">arange<\/span><span class=\"p\">(<\/span><span class=\"mi\">10<\/span><span class=\"p\">))<\/span>\r\n    <span class=\"n\">plt<\/span><span class=\"o\">.<\/span><span class=\"n\">yticks<\/span><span class=\"p\">(<\/span><span class=\"n\">np<\/span><span class=\"o\">.<\/span><span class=\"n\">arange<\/span><span class=\"p\">(<\/span><span class=\"mi\">10<\/span><span class=\"p\">))<\/span>\r\n\r\n    <span class=\"n\">plt<\/span><span class=\"o\">.<\/span><span class=\"n\">tight_layout<\/span><span class=\"p\">()<\/span>\r\n    <span class=\"n\">plt<\/span><span class=\"o\">.<\/span><span class=\"n\">show<\/span><span class=\"p\">()<\/span>\r\n\r\n    <span class=\"k\">return<\/span> <span class=\"n\">cm_norm<\/span>\r\n<\/pre>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing text_cell rendered\">\n<div class=\"inner_cell\">\n<div class=\"text_cell_render border-box-sizing rendered_html\">\n<p>\u307e\u305a\u306f\u3001\u8aa4\u5206\u985e\u3092\u8d77\u3053\u3057\u3066\u3057\u307e\u3063\u305f\u753b\u50cf\u3092\u8868\u793a\u3055\u305b\u307e\u3059\u3002<\/p>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing code_cell rendered\">\n<div class=\"input\">\n<div class=\"inner_cell\">\n<div class=\"input_area\">\n<div class=\"highlight hl-python\">\n<pre><span><\/span><span class=\"nb\">print<\/span><span class=\"p\">(<\/span><span class=\"s2\">\"=== \u8aa4\u5206\u985e\u753b\u50cf\u306e\u8868\u793a ===\"<\/span><span class=\"p\">)<\/span>\r\n\r\n<span class=\"n\">show_misclassified<\/span><span class=\"p\">(<\/span><span class=\"n\">X_test<\/span><span class=\"p\">,<\/span> <span class=\"n\">y_test<\/span><span class=\"p\">,<\/span> <span class=\"n\">y_pred<\/span><span class=\"p\">,<\/span> <span class=\"n\">num<\/span><span class=\"o\">=<\/span><span class=\"mi\">5<\/span><span class=\"p\">)<\/span>\r\n<\/pre>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"output_wrapper\">\n<div class=\"output\">\n<div class=\"output_area\">\n<div class=\"prompt\"><\/div>\n<div class=\"output_subarea output_stream output_stdout output_text\">\n<pre>=== \u8aa4\u5206\u985e\u753b\u50cf\u306e\u8868\u793a ===\r\n<\/pre>\n<\/div>\n<\/div>\n<div class=\"output_area\">\n<div class=\"prompt\"><\/div>\n<div class=\"output_png output_subarea\"><img decoding=\"async\" alt=\"No description has been provided for this image\" src=\"data:image\/png;base64,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\" \/><\/div>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing text_cell rendered\">\n<div class=\"inner_cell\">\n<div class=\"text_cell_render border-box-sizing rendered_html\">\n<p>\u7d9a\u3044\u3066\u3001\u6df7\u540c\u884c\u5217\u3092\u30d2\u30fc\u30c8\u30de\u30c3\u30d7\u3068\u3057\u3066\u8868\u793a\u3057\u307e\u3059\u3002<\/p>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing code_cell rendered\">\n<div class=\"input\">\n<div class=\"inner_cell\">\n<div class=\"input_area\">\n<div class=\"highlight hl-python\">\n<pre><span><\/span><span class=\"nb\">print<\/span><span class=\"p\">(<\/span><span class=\"s2\">\"=== \u6df7\u540c\u884c\u5217\u306e\u8868\u793a ===\"<\/span><span class=\"p\">)<\/span>\r\n\r\n<span class=\"n\">cm_norm<\/span> <span class=\"o\">=<\/span> <span class=\"n\">plot_normalized_confusion_matrix_heatmap<\/span><span class=\"p\">(<\/span><span class=\"n\">y_test<\/span><span class=\"p\">,<\/span> <span class=\"n\">y_pred<\/span><span class=\"p\">)<\/span>\r\n<\/pre>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"output_wrapper\">\n<div class=\"output\">\n<div class=\"output_area\">\n<div class=\"prompt\"><\/div>\n<div class=\"output_subarea output_stream output_stdout output_text\">\n<pre>=== \u6df7\u540c\u884c\u5217\u306e\u8868\u793a ===\r\n<\/pre>\n<\/div>\n<\/div>\n<div class=\"output_area\">\n<div class=\"prompt\"><\/div>\n<div class=\"output_png output_subarea\"><img decoding=\"async\" alt=\"No description has been provided for this image\" src=\"data:image\/png;base64,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\" \/><\/div>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing text_cell rendered\">\n<div class=\"inner_cell\">\n<div class=\"text_cell_render border-box-sizing rendered_html\">\n<p>\u3053\u306e\u6df7\u540c\u884c\u5217\uff08\u30d2\u30fc\u30c8\u30de\u30c3\u30d7\uff09\u306f\u3001\u7e26\u8ef8\u304c\u6b63\u89e3\u30e9\u30d9\u30eb\u3001\u6a2a\u8ef8\u304c\u4e88\u6e2c\u30e9\u30d9\u30eb\u3092\u8868\u3057\u3066\u3044\u307e\u3059\u3002\u3057\u305f\u304c\u3063\u3066\u3001\u5bfe\u89d2\u6210\u5206\u306e\u5024\u304c\u9ad8\u3044\u307b\u3069\u3001\u305d\u306e\u6570\u5b57\u3092\u6b63\u3057\u304f\u5206\u985e\u3067\u304d\u305f\u5272\u5408\uff08\u6b63\u89e3\u7387\uff09\u304c\u9ad8\u3044\u3053\u3068\u3092\u610f\u5473\u3057\u307e\u3059\u3002\u3053\u306e\u6df7\u540c\u884c\u5217\u306e\u7d50\u679c\u304b\u3089\u3001NBMF\u306f\u5f62\u72b6\u304c\u30b7\u30f3\u30d7\u30eb\u306a\u300c0\u300d\u3084\u300c1\u300d\u306e\u5206\u985e\u3092\u5f97\u610f\u3068\u3059\u308b\u4e00\u65b9\u3067\u3001\u300c2\u300d\u300c5\u300d\u300c9\u300d\u306e\u3088\u3046\u306a\u8907\u96d1\u306a\u5f62\u72b6\u3092\u6301\u3064\u6570\u5b57\u306e\u5206\u985e\u306b\u306f\u8ab2\u984c\u304c\u3042\u308b\u3053\u3068\u304c\u8aad\u307f\u53d6\u308c\u307e\u3059\u3002\u5b9f\u969b\u306b\u8aa4\u5206\u985e\u3055\u308c\u305f\u753b\u50cf\u3092\u89b3\u5bdf\u3059\u308b\u3068\u3001\u3053\u308c\u3089\u306e\u6570\u5b57\u306f\u4ed6\u306e\u6570\u5b57\u3068\u985e\u4f3c\u3057\u305f\u7279\u5fb4\u3092\u4f75\u305b\u6301\u3063\u3066\u304a\u308a\u3001\u305d\u308c\u304c\u30e2\u30c7\u30eb\u306e\u5224\u65ad\u3092\u8ff7\u308f\u305b\u3066\u3044\u308b\u3053\u3068\u304c\u78ba\u8a8d\u3067\u304d\u307e\u3057\u305f\u3002<\/p>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing text_cell rendered\">\n<div class=\"inner_cell\">\n<div class=\"text_cell_render border-box-sizing rendered_html\">\n<h2><span class=\"ez-toc-section\" id=\"%E3%81%BE%E3%81%A8%E3%82%81\"><\/span>\u307e\u3068\u3081<span class=\"ez-toc-section-end\"><\/span><\/h2>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing text_cell rendered\">\n<div class=\"inner_cell\">\n<div class=\"text_cell_render border-box-sizing rendered_html\">\n<p>\u672c\u8a18\u4e8b\u3067\u306f\u3001\u5143\u8ad6\u6587\u306e\u975e\u8ca0\u4e8c\u5024\u884c\u5217\u56e0\u5b50\u5206\u89e3\uff08NBMF\uff09\u30a2\u30eb\u30b4\u30ea\u30ba\u30e0\u3092\u5b9f\u88c5\u3057\u3001\u624b\u66f8\u304d\u6570\u5b57\u5206\u985e\u30bf\u30b9\u30af\u306b\u304a\u3044\u3066\u5168\u7d50\u5408\u30cb\u30e5\u30fc\u30e9\u30eb\u30cd\u30c3\u30c8\u30ef\u30fc\u30af\uff08FCNN\uff09\u3068\u306e\u6027\u80fd\u6bd4\u8f03\u3092\u884c\u3044\u307e\u3057\u305f\u3002 \u5143\u8ad6\u6587\u3068\u540c\u69d8\u306e\u6761\u4ef6\u3067\u691c\u8a3c\u3057\u305f\u7d50\u679c\u30013\u3064\u306e\u5b9f\u9a13\u3059\u3079\u3066\u306b\u304a\u3044\u3066\u8ad6\u6587\u306e\u5831\u544a\u306b\u8fd1\u3044\u6027\u80fd\u3092\u518d\u73fe\u3059\u308b\u3053\u3068\u304c\u3067\u304d\u307e\u3057\u305f\u3002\u307e\u305f\u3001\u5b66\u7fd2\u7d50\u679c\u306e\u53ef\u8996\u5316\u3092\u901a\u3057\u3066\u3001\u30e2\u30c7\u30eb\u304c\u6570\u5b57\u306e\u7279\u5fb4\u3092\u3069\u306e\u3088\u3046\u306b\u6349\u3048\u3066\u3044\u308b\u304b\u3092\u8996\u899a\u7684\u306b\u78ba\u8a8d\u3059\u308b\u3053\u3068\u304c\u3067\u304d\u307e\u3057\u305f\u3002<\/p>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing text_cell rendered\">\n<div class=\"inner_cell\">\n<div class=\"text_cell_render border-box-sizing rendered_html\">\n<h2><span class=\"ez-toc-section\" id=\"%E3%81%82%E3%81%A8%E3%81%8C%E3%81%8D\"><\/span>\u3042\u3068\u304c\u304d<span class=\"ez-toc-section-end\"><\/span><\/h2>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing text_cell rendered\">\n<div class=\"inner_cell\">\n<div class=\"text_cell_render border-box-sizing rendered_html\">\n<p>\u975e\u8ca0\u4e8c\u5024\u884c\u5217\u56e0\u5b50\u5206\u89e3\uff08NBMF\uff09\u3068\u3044\u3046\u99b4\u67d3\u307f\u306e\u306a\u304b\u3063\u305f\u624b\u6cd5\u3092\u8ad6\u6587\u304b\u3089\u5b66\u3073\u3001\u305d\u308c\u3092\u5b9f\u969b\u306b\u91cf\u5b50\u30a2\u30cb\u30fc\u30ea\u30f3\u30b0\uff08\u30b7\u30df\u30e5\u30ec\u30fc\u30bf\uff09\u3067\u5b9f\u88c5\u3059\u308b\u3068\u3044\u3046\u4e00\u9023\u306e\u30d7\u30ed\u30bb\u30b9\u306f\u975e\u5e38\u306b\u52c9\u5f37\u306b\u306a\u308a\u307e\u3057\u305f\u3002\u4eca\u5f8c\u306f\u3001\u8aa4\u5206\u985e\u304c\u591a\u304b\u3063\u305f\u8907\u96d1\u306a\u6570\u5b57\u306b\u5bfe\u3059\u308b\u8a8d\u8b58\u7cbe\u5ea6\u306e\u5411\u4e0a\u3084\u3001\u5b9f\u6a5f\u306e\u30a2\u30cb\u30fc\u30ea\u30f3\u30b0\u30de\u30b7\u30f3\u3092\u7528\u3044\u305f\u691c\u8a3c\u306b\u3082\u6311\u6226\u3057\u3066\u307f\u305f\u3044\u3068\u601d\u3044\u307e\u3059\u3002<\/p>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing text_cell rendered\">\n<div class=\"inner_cell\">\n<div class=\"text_cell_render border-box-sizing rendered_html\">\n<h3><span class=\"ez-toc-section\" id=\"%E6%9C%AC%E8%A8%98%E4%BA%8B%E3%81%AE%E4%BD%9C%E6%88%90%E8%80%85\"><\/span>\u672c\u8a18\u4e8b\u306e\u4f5c\u6210\u8005<span class=\"ez-toc-section-end\"><\/span><\/h3>\n<\/div>\n<\/div>\n<\/div>\n<div class=\"cell border-box-sizing text_cell rendered\">\n<div class=\"inner_cell\">\n<div class=\"text_cell_render border-box-sizing rendered_html\">\n<p>\u5c71\u7530\u7adc\u96c5<\/p>\n<\/div>\n<\/div>\n<\/div>\n","protected":false},"excerpt":{"rendered":"<p>\u6982\u8981 \u89e3\u8aac\u8a18\u4e8b\u300c\u91cf\u5b50\u30a2\u30cb\u30fc\u30ea\u30f3\u30b0\u3092\u6d3b\u7528\u3057\u305f\u975e\u8ca0\u4e8c\u5024\u884c\u5217\u56e0\u5b50\u5206\u89e3\u306b\u3088\u308b\u753b\u50cf\u5206\u985e\u300d\u3067\u306f\u3001\u975e\u8ca0\u4e8c\u5024\u884c\u5217\u56e0\u5b50\u5206\u89e3\u3092\u5fdc\u7528\u3057\u305f\u591a\u30af\u30e9\u30b9\u753b\u50cf\u5206\u985e\u30e2\u30c7\u30eb\u3092\u63d0\u6848\u3057, \u624b\u66f8\u304d\u6570\u5b57\u753b\u50cf\u3092\u5206\u985e\u3059\u308b\u3068\u3044\u3046\u554f\u984c\u306b\u5bfe\u3057\u3066\u63d0\u6848\u30e2\u30c7\u30eb\u3068\u5168\u7d50\u5408\u30cb\u30e5\u30fc\u30e9\u30eb\u30cd [&hellip;]<\/p>\n","protected":false},"author":20,"featured_media":8650,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[4],"tags":[60,94,109,122,129],"class_list":["post-8643","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-hands-on","tag-60","tag-94","tag-109","tag-122","tag-129"],"yoast_head":"<!-- This site is optimized with the Yoast SEO plugin v27.4 - https:\/\/yoast.com\/product\/yoast-seo-wordpress\/ -->\n<title>\u3010\u5b9f\u8df5\u7de8\u3011\u91cf\u5b50\u30a2\u30cb\u30fc\u30ea\u30f3\u30b0\u3092\u6d3b\u7528\u3057\u305f\u975e\u8ca0\u4e8c\u5024\u884c\u5217\u56e0\u5b50\u5206\u89e3\u306b\u3088\u308b\u753b\u50cf\u5206\u985e - T-QARD Harbor<\/title>\n<meta name=\"description\" content=\"\u89e3\u8aac\u8a18\u4e8b\u300c\u91cf\u5b50\u30a2\u30cb\u30fc\u30ea\u30f3\u30b0\u3092\u6d3b\u7528\u3057\u305f\u975e\u8ca0\u4e8c\u5024\u884c\u5217\u56e0\u5b50\u5206\u89e3\u306b\u3088\u308b\u753b\u50cf\u5206\u985e\u300d\u3067\u306f\u3001\u975e\u8ca0\u4e8c\u5024\u884c\u5217\u56e0\u5b50\u5206\u89e3\u3092\u5fdc\u7528\u3057\u305f\u591a\u30af\u30e9\u30b9\u753b\u50cf\u5206\u985e\u30e2\u30c7\u30eb\u3092\u63d0\u6848\u3057, \u624b\u66f8\u304d\u6570\u5b57\u753b\u50cf\u3092\u5206\u985e\u3059\u308b\u3068\u3044\u3046\u554f\u984c\u306b\u5bfe\u3057\u3066\u63d0\u6848\u30e2\u30c7\u30eb\u3068\u5168\u7d50\u5408\u30cb\u30e5\u30fc\u30e9\u30eb\u30cd\u30c3\u30c8\u30ef\u30fc\u30af\u3067\u305d\u306e\u6027\u80fd\u3092\u6bd4\u8f03\u3059\u308b\u8ad6\u6587\u3092\u7d39\u4ecb\u3057\u307e\u3057\u305f\u3002\u672c\u8a18\u4e8b\u3067\u306f\u3001\u305d\u306e\u30a2\u30eb\u30b4\u30ea\u30ba\u30e0\u3092\u5b9f\u88c5\u3057\u3001\u5143\u8ad6\u6587\u306e\u518d\u73fe\u5b9f\u9a13\u3092\u884c\u3044\u307e\u3059\u3002\" \/>\n<meta name=\"robots\" content=\"index, follow, max-snippet:-1, max-image-preview:large, max-video-preview:-1\" \/>\n<link rel=\"canonical\" href=\"https:\/\/qard.is.tohoku.ac.jp\/T-Wave\/2026\/02\/13\/\u3010\u5b9f\u8df5\u7de8\u3011\u91cf\u5b50\u30a2\u30cb\u30fc\u30ea\u30f3\u30b0\u3092\u6d3b\u7528\u3057\u305f\u975e\u8ca0\u4e8c\u5024\/\" \/>\n<meta property=\"og:locale\" content=\"ja_JP\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"\u3010\u5b9f\u8df5\u7de8\u3011\u91cf\u5b50\u30a2\u30cb\u30fc\u30ea\u30f3\u30b0\u3092\u6d3b\u7528\u3057\u305f\u975e\u8ca0\u4e8c\u5024\u884c\u5217\u56e0\u5b50\u5206\u89e3\u306b\u3088\u308b\u753b\u50cf\u5206\u985e - T-QARD Harbor\" \/>\n<meta property=\"og:description\" content=\"\u89e3\u8aac\u8a18\u4e8b\u300c\u91cf\u5b50\u30a2\u30cb\u30fc\u30ea\u30f3\u30b0\u3092\u6d3b\u7528\u3057\u305f\u975e\u8ca0\u4e8c\u5024\u884c\u5217\u56e0\u5b50\u5206\u89e3\u306b\u3088\u308b\u753b\u50cf\u5206\u985e\u300d\u3067\u306f\u3001\u975e\u8ca0\u4e8c\u5024\u884c\u5217\u56e0\u5b50\u5206\u89e3\u3092\u5fdc\u7528\u3057\u305f\u591a\u30af\u30e9\u30b9\u753b\u50cf\u5206\u985e\u30e2\u30c7\u30eb\u3092\u63d0\u6848\u3057, \u624b\u66f8\u304d\u6570\u5b57\u753b\u50cf\u3092\u5206\u985e\u3059\u308b\u3068\u3044\u3046\u554f\u984c\u306b\u5bfe\u3057\u3066\u63d0\u6848\u30e2\u30c7\u30eb\u3068\u5168\u7d50\u5408\u30cb\u30e5\u30fc\u30e9\u30eb\u30cd\u30c3\u30c8\u30ef\u30fc\u30af\u3067\u305d\u306e\u6027\u80fd\u3092\u6bd4\u8f03\u3059\u308b\u8ad6\u6587\u3092\u7d39\u4ecb\u3057\u307e\u3057\u305f\u3002\u672c\u8a18\u4e8b\u3067\u306f\u3001\u305d\u306e\u30a2\u30eb\u30b4\u30ea\u30ba\u30e0\u3092\u5b9f\u88c5\u3057\u3001\u5143\u8ad6\u6587\u306e\u518d\u73fe\u5b9f\u9a13\u3092\u884c\u3044\u307e\u3059\u3002\" \/>\n<meta property=\"og:url\" content=\"https:\/\/qard.is.tohoku.ac.jp\/T-Wave\/2026\/02\/13\/\u3010\u5b9f\u8df5\u7de8\u3011\u91cf\u5b50\u30a2\u30cb\u30fc\u30ea\u30f3\u30b0\u3092\u6d3b\u7528\u3057\u305f\u975e\u8ca0\u4e8c\u5024\/\" \/>\n<meta property=\"og:site_name\" content=\"T-QARD Harbor\" \/>\n<meta property=\"article:published_time\" content=\"2026-02-13T01:12:46+00:00\" \/>\n<meta property=\"og:image\" content=\"https:\/\/qard.is.tohoku.ac.jp\/T-Wave\/wp-content\/uploads\/2026\/02\/b7cd83984219402b84f667ac12783404.png\" \/>\n\t<meta property=\"og:image:width\" content=\"1500\" \/>\n\t<meta property=\"og:image:height\" content=\"1213\" \/>\n\t<meta property=\"og:image:type\" content=\"image\/png\" \/>\n<meta name=\"author\" content=\"Ryuga Yamada\" \/>\n<meta name=\"twitter:card\" content=\"summary_large_image\" \/>\n<meta name=\"twitter:label1\" content=\"\u57f7\u7b46\u8005\" \/>\n\t<meta name=\"twitter:data1\" content=\"Ryuga Yamada\" \/>\n\t<meta name=\"twitter:label2\" content=\"\u63a8\u5b9a\u8aad\u307f\u53d6\u308a\u6642\u9593\" \/>\n\t<meta name=\"twitter:data2\" content=\"161\u5206\" \/>\n<script type=\"application\/ld+json\" class=\"yoast-schema-graph\">{\"@context\":\"https:\\\/\\\/schema.org\",\"@graph\":[{\"@type\":\"Article\",\"@id\":\"https:\\\/\\\/qard.is.tohoku.ac.jp\\\/T-Wave\\\/2026\\\/02\\\/13\\\/%e3%80%90%e5%ae%9f%e8%b7%b5%e7%b7%a8%e3%80%91%e9%87%8f%e5%ad%90%e3%82%a2%e3%83%8b%e3%83%bc%e3%83%aa%e3%83%b3%e3%82%b0%e3%82%92%e6%b4%bb%e7%94%a8%e3%81%97%e3%81%9f%e9%9d%9e%e8%b2%a0%e4%ba%8c%e5%80%a4\\\/#article\",\"isPartOf\":{\"@id\":\"https:\\\/\\\/qard.is.tohoku.ac.jp\\\/T-Wave\\\/2026\\\/02\\\/13\\\/%e3%80%90%e5%ae%9f%e8%b7%b5%e7%b7%a8%e3%80%91%e9%87%8f%e5%ad%90%e3%82%a2%e3%83%8b%e3%83%bc%e3%83%aa%e3%83%b3%e3%82%b0%e3%82%92%e6%b4%bb%e7%94%a8%e3%81%97%e3%81%9f%e9%9d%9e%e8%b2%a0%e4%ba%8c%e5%80%a4\\\/\"},\"author\":{\"name\":\"Ryuga Yamada\",\"@id\":\"https:\\\/\\\/qard.is.tohoku.ac.jp\\\/T-Wave\\\/#\\\/schema\\\/person\\\/be616b43e2d5a6ad431f358bc5bd34f4\"},\"headline\":\"\u3010\u5b9f\u8df5\u7de8\u3011\u91cf\u5b50\u30a2\u30cb\u30fc\u30ea\u30f3\u30b0\u3092\u6d3b\u7528\u3057\u305f\u975e\u8ca0\u4e8c\u5024\u884c\u5217\u56e0\u5b50\u5206\u89e3\u306b\u3088\u308b\u753b\u50cf\u5206\u985e\",\"datePublished\":\"2026-02-13T01:12:46+00:00\",\"mainEntityOfPage\":{\"@id\":\"https:\\\/\\\/qard.is.tohoku.ac.jp\\\/T-Wave\\\/2026\\\/02\\\/13\\\/%e3%80%90%e5%ae%9f%e8%b7%b5%e7%b7%a8%e3%80%91%e9%87%8f%e5%ad%90%e3%82%a2%e3%83%8b%e3%83%bc%e3%83%aa%e3%83%b3%e3%82%b0%e3%82%92%e6%b4%bb%e7%94%a8%e3%81%97%e3%81%9f%e9%9d%9e%e8%b2%a0%e4%ba%8c%e5%80%a4\\\/\"},\"wordCount\":383,\"commentCount\":0,\"image\":{\"@id\":\"https:\\\/\\\/qard.is.tohoku.ac.jp\\\/T-Wave\\\/2026\\\/02\\\/13\\\/%e3%80%90%e5%ae%9f%e8%b7%b5%e7%b7%a8%e3%80%91%e9%87%8f%e5%ad%90%e3%82%a2%e3%83%8b%e3%83%bc%e3%83%aa%e3%83%b3%e3%82%b0%e3%82%92%e6%b4%bb%e7%94%a8%e3%81%97%e3%81%9f%e9%9d%9e%e8%b2%a0%e4%ba%8c%e5%80%a4\\\/#primaryimage\"},\"thumbnailUrl\":\"https:\\\/\\\/qard.is.tohoku.ac.jp\\\/T-Wave\\\/wp-content\\\/uploads\\\/2026\\\/02\\\/b7cd83984219402b84f667ac12783404.png\",\"keywords\":[\"\u30cb\u30e5\u30fc\u30e9\u30eb\u30cd\u30c3\u30c8\u30ef\u30fc\u30af\",\"\u5b9f\u8df5\u8a18\u4e8b\",\"\u753b\u50cf\u5206\u985e\",\"\u91cf\u5b50\u30a2\u30cb\u30fc\u30ea\u30f3\u30b0\",\"\u975e\u8ca0\u4e8c\u5024\u884c\u5217\u56e0\u5b50\u5206\u89e3\"],\"articleSection\":[\"\u5b9f\u8df5\u8a18\u4e8b\"],\"inLanguage\":\"ja\",\"potentialAction\":[{\"@type\":\"CommentAction\",\"name\":\"Comment\",\"target\":[\"https:\\\/\\\/qard.is.tohoku.ac.jp\\\/T-Wave\\\/2026\\\/02\\\/13\\\/%e3%80%90%e5%ae%9f%e8%b7%b5%e7%b7%a8%e3%80%91%e9%87%8f%e5%ad%90%e3%82%a2%e3%83%8b%e3%83%bc%e3%83%aa%e3%83%b3%e3%82%b0%e3%82%92%e6%b4%bb%e7%94%a8%e3%81%97%e3%81%9f%e9%9d%9e%e8%b2%a0%e4%ba%8c%e5%80%a4\\\/#respond\"]}]},{\"@type\":\"WebPage\",\"@id\":\"https:\\\/\\\/qard.is.tohoku.ac.jp\\\/T-Wave\\\/2026\\\/02\\\/13\\\/%e3%80%90%e5%ae%9f%e8%b7%b5%e7%b7%a8%e3%80%91%e9%87%8f%e5%ad%90%e3%82%a2%e3%83%8b%e3%83%bc%e3%83%aa%e3%83%b3%e3%82%b0%e3%82%92%e6%b4%bb%e7%94%a8%e3%81%97%e3%81%9f%e9%9d%9e%e8%b2%a0%e4%ba%8c%e5%80%a4\\\/\",\"url\":\"https:\\\/\\\/qard.is.tohoku.ac.jp\\\/T-Wave\\\/2026\\\/02\\\/13\\\/%e3%80%90%e5%ae%9f%e8%b7%b5%e7%b7%a8%e3%80%91%e9%87%8f%e5%ad%90%e3%82%a2%e3%83%8b%e3%83%bc%e3%83%aa%e3%83%b3%e3%82%b0%e3%82%92%e6%b4%bb%e7%94%a8%e3%81%97%e3%81%9f%e9%9d%9e%e8%b2%a0%e4%ba%8c%e5%80%a4\\\/\",\"name\":\"\u3010\u5b9f\u8df5\u7de8\u3011\u91cf\u5b50\u30a2\u30cb\u30fc\u30ea\u30f3\u30b0\u3092\u6d3b\u7528\u3057\u305f\u975e\u8ca0\u4e8c\u5024\u884c\u5217\u56e0\u5b50\u5206\u89e3\u306b\u3088\u308b\u753b\u50cf\u5206\u985e - T-QARD Harbor\",\"isPartOf\":{\"@id\":\"https:\\\/\\\/qard.is.tohoku.ac.jp\\\/T-Wave\\\/#website\"},\"primaryImageOfPage\":{\"@id\":\"https:\\\/\\\/qard.is.tohoku.ac.jp\\\/T-Wave\\\/2026\\\/02\\\/13\\\/%e3%80%90%e5%ae%9f%e8%b7%b5%e7%b7%a8%e3%80%91%e9%87%8f%e5%ad%90%e3%82%a2%e3%83%8b%e3%83%bc%e3%83%aa%e3%83%b3%e3%82%b0%e3%82%92%e6%b4%bb%e7%94%a8%e3%81%97%e3%81%9f%e9%9d%9e%e8%b2%a0%e4%ba%8c%e5%80%a4\\\/#primaryimage\"},\"image\":{\"@id\":\"https:\\\/\\\/qard.is.tohoku.ac.jp\\\/T-Wave\\\/2026\\\/02\\\/13\\\/%e3%80%90%e5%ae%9f%e8%b7%b5%e7%b7%a8%e3%80%91%e9%87%8f%e5%ad%90%e3%82%a2%e3%83%8b%e3%83%bc%e3%83%aa%e3%83%b3%e3%82%b0%e3%82%92%e6%b4%bb%e7%94%a8%e3%81%97%e3%81%9f%e9%9d%9e%e8%b2%a0%e4%ba%8c%e5%80%a4\\\/#primaryimage\"},\"thumbnailUrl\":\"https:\\\/\\\/qard.is.tohoku.ac.jp\\\/T-Wave\\\/wp-content\\\/uploads\\\/2026\\\/02\\\/b7cd83984219402b84f667ac12783404.png\",\"datePublished\":\"2026-02-13T01:12:46+00:00\",\"author\":{\"@id\":\"https:\\\/\\\/qard.is.tohoku.ac.jp\\\/T-Wave\\\/#\\\/schema\\\/person\\\/be616b43e2d5a6ad431f358bc5bd34f4\"},\"description\":\"\u89e3\u8aac\u8a18\u4e8b\u300c\u91cf\u5b50\u30a2\u30cb\u30fc\u30ea\u30f3\u30b0\u3092\u6d3b\u7528\u3057\u305f\u975e\u8ca0\u4e8c\u5024\u884c\u5217\u56e0\u5b50\u5206\u89e3\u306b\u3088\u308b\u753b\u50cf\u5206\u985e\u300d\u3067\u306f\u3001\u975e\u8ca0\u4e8c\u5024\u884c\u5217\u56e0\u5b50\u5206\u89e3\u3092\u5fdc\u7528\u3057\u305f\u591a\u30af\u30e9\u30b9\u753b\u50cf\u5206\u985e\u30e2\u30c7\u30eb\u3092\u63d0\u6848\u3057, \u624b\u66f8\u304d\u6570\u5b57\u753b\u50cf\u3092\u5206\u985e\u3059\u308b\u3068\u3044\u3046\u554f\u984c\u306b\u5bfe\u3057\u3066\u63d0\u6848\u30e2\u30c7\u30eb\u3068\u5168\u7d50\u5408\u30cb\u30e5\u30fc\u30e9\u30eb\u30cd\u30c3\u30c8\u30ef\u30fc\u30af\u3067\u305d\u306e\u6027\u80fd\u3092\u6bd4\u8f03\u3059\u308b\u8ad6\u6587\u3092\u7d39\u4ecb\u3057\u307e\u3057\u305f\u3002\u672c\u8a18\u4e8b\u3067\u306f\u3001\u305d\u306e\u30a2\u30eb\u30b4\u30ea\u30ba\u30e0\u3092\u5b9f\u88c5\u3057\u3001\u5143\u8ad6\u6587\u306e\u518d\u73fe\u5b9f\u9a13\u3092\u884c\u3044\u307e\u3059\u3002\",\"breadcrumb\":{\"@id\":\"https:\\\/\\\/qard.is.tohoku.ac.jp\\\/T-Wave\\\/2026\\\/02\\\/13\\\/%e3%80%90%e5%ae%9f%e8%b7%b5%e7%b7%a8%e3%80%91%e9%87%8f%e5%ad%90%e3%82%a2%e3%83%8b%e3%83%bc%e3%83%aa%e3%83%b3%e3%82%b0%e3%82%92%e6%b4%bb%e7%94%a8%e3%81%97%e3%81%9f%e9%9d%9e%e8%b2%a0%e4%ba%8c%e5%80%a4\\\/#breadcrumb\"},\"inLanguage\":\"ja\",\"potentialAction\":[{\"@type\":\"ReadAction\",\"target\":[\"https:\\\/\\\/qard.is.tohoku.ac.jp\\\/T-Wave\\\/2026\\\/02\\\/13\\\/%e3%80%90%e5%ae%9f%e8%b7%b5%e7%b7%a8%e3%80%91%e9%87%8f%e5%ad%90%e3%82%a2%e3%83%8b%e3%83%bc%e3%83%aa%e3%83%b3%e3%82%b0%e3%82%92%e6%b4%bb%e7%94%a8%e3%81%97%e3%81%9f%e9%9d%9e%e8%b2%a0%e4%ba%8c%e5%80%a4\\\/\"]}]},{\"@type\":\"ImageObject\",\"inLanguage\":\"ja\",\"@id\":\"https:\\\/\\\/qard.is.tohoku.ac.jp\\\/T-Wave\\\/2026\\\/02\\\/13\\\/%e3%80%90%e5%ae%9f%e8%b7%b5%e7%b7%a8%e3%80%91%e9%87%8f%e5%ad%90%e3%82%a2%e3%83%8b%e3%83%bc%e3%83%aa%e3%83%b3%e3%82%b0%e3%82%92%e6%b4%bb%e7%94%a8%e3%81%97%e3%81%9f%e9%9d%9e%e8%b2%a0%e4%ba%8c%e5%80%a4\\\/#primaryimage\",\"url\":\"https:\\\/\\\/qard.is.tohoku.ac.jp\\\/T-Wave\\\/wp-content\\\/uploads\\\/2026\\\/02\\\/b7cd83984219402b84f667ac12783404.png\",\"contentUrl\":\"https:\\\/\\\/qard.is.tohoku.ac.jp\\\/T-Wave\\\/wp-content\\\/uploads\\\/2026\\\/02\\\/b7cd83984219402b84f667ac12783404.png\",\"width\":1500,\"height\":1213},{\"@type\":\"BreadcrumbList\",\"@id\":\"https:\\\/\\\/qard.is.tohoku.ac.jp\\\/T-Wave\\\/2026\\\/02\\\/13\\\/%e3%80%90%e5%ae%9f%e8%b7%b5%e7%b7%a8%e3%80%91%e9%87%8f%e5%ad%90%e3%82%a2%e3%83%8b%e3%83%bc%e3%83%aa%e3%83%b3%e3%82%b0%e3%82%92%e6%b4%bb%e7%94%a8%e3%81%97%e3%81%9f%e9%9d%9e%e8%b2%a0%e4%ba%8c%e5%80%a4\\\/#breadcrumb\",\"itemListElement\":[{\"@type\":\"ListItem\",\"position\":1,\"name\":\"\u30db\u30fc\u30e0\",\"item\":\"https:\\\/\\\/qard.is.tohoku.ac.jp\\\/T-Wave\\\/\"},{\"@type\":\"ListItem\",\"position\":2,\"name\":\"\u3010\u5b9f\u8df5\u7de8\u3011\u91cf\u5b50\u30a2\u30cb\u30fc\u30ea\u30f3\u30b0\u3092\u6d3b\u7528\u3057\u305f\u975e\u8ca0\u4e8c\u5024\u884c\u5217\u56e0\u5b50\u5206\u89e3\u306b\u3088\u308b\u753b\u50cf\u5206\u985e\"}]},{\"@type\":\"WebSite\",\"@id\":\"https:\\\/\\\/qard.is.tohoku.ac.jp\\\/T-Wave\\\/#website\",\"url\":\"https:\\\/\\\/qard.is.tohoku.ac.jp\\\/T-Wave\\\/\",\"name\":\"T-QARD Harbor\",\"description\":\"T-QARD Harbor\u306f\u6771\u5317\u5927\u5b66\u91cf\u5b50\u30a2\u30d7\u30ea\u30b1\u30fc\u30b7\u30e7\u30f3\u7814\u7a76\u958b\u767a\u30bb\u30f3\u30bf\u30fc\u5b66\u751f\u30c1\u30fc\u30e0\u300cT-QARD Crews\u300d\u304c\u904b\u55b6\u3059\u308b\u3001 \u6570\u7406\u60c5\u5831\u7d71\u8a08\u3001\u91cf\u5b50\u60c5\u5831\u3001\u6700\u9069\u5316\u3001\u6a5f\u68b0\u5b66\u7fd2\u5206\u91ce\u306e\u60c5\u5831\u3092\u63d0\u4f9b\u3059\u308bWeb\u30b5\u30a4\u30c8\u3067\u3059\",\"potentialAction\":[{\"@type\":\"SearchAction\",\"target\":{\"@type\":\"EntryPoint\",\"urlTemplate\":\"https:\\\/\\\/qard.is.tohoku.ac.jp\\\/T-Wave\\\/?s={search_term_string}\"},\"query-input\":{\"@type\":\"PropertyValueSpecification\",\"valueRequired\":true,\"valueName\":\"search_term_string\"}}],\"inLanguage\":\"ja\"},{\"@type\":\"Person\",\"@id\":\"https:\\\/\\\/qard.is.tohoku.ac.jp\\\/T-Wave\\\/#\\\/schema\\\/person\\\/be616b43e2d5a6ad431f358bc5bd34f4\",\"name\":\"Ryuga Yamada\",\"image\":{\"@type\":\"ImageObject\",\"inLanguage\":\"ja\",\"@id\":\"https:\\\/\\\/secure.gravatar.com\\\/avatar\\\/6f1965735667140ac51f351536e86b546a2ca3406c60bbf23e27ae7de1091db1?s=96&d=mm&r=g\",\"url\":\"https:\\\/\\\/secure.gravatar.com\\\/avatar\\\/6f1965735667140ac51f351536e86b546a2ca3406c60bbf23e27ae7de1091db1?s=96&d=mm&r=g\",\"contentUrl\":\"https:\\\/\\\/secure.gravatar.com\\\/avatar\\\/6f1965735667140ac51f351536e86b546a2ca3406c60bbf23e27ae7de1091db1?s=96&d=mm&r=g\",\"caption\":\"Ryuga Yamada\"},\"description\":\"\u6771\u5317\u5927\u5b66 \u5de5\u5b66\u90e8 \u5b66\u58eb\u8ab2\u7a0b\",\"url\":\"https:\\\/\\\/qard.is.tohoku.ac.jp\\\/T-Wave\\\/author\\\/ryuga-yamada\\\/\"}]}<\/script>\n<!-- \/ Yoast SEO plugin. -->","yoast_head_json":{"title":"\u3010\u5b9f\u8df5\u7de8\u3011\u91cf\u5b50\u30a2\u30cb\u30fc\u30ea\u30f3\u30b0\u3092\u6d3b\u7528\u3057\u305f\u975e\u8ca0\u4e8c\u5024\u884c\u5217\u56e0\u5b50\u5206\u89e3\u306b\u3088\u308b\u753b\u50cf\u5206\u985e - T-QARD Harbor","description":"\u89e3\u8aac\u8a18\u4e8b\u300c\u91cf\u5b50\u30a2\u30cb\u30fc\u30ea\u30f3\u30b0\u3092\u6d3b\u7528\u3057\u305f\u975e\u8ca0\u4e8c\u5024\u884c\u5217\u56e0\u5b50\u5206\u89e3\u306b\u3088\u308b\u753b\u50cf\u5206\u985e\u300d\u3067\u306f\u3001\u975e\u8ca0\u4e8c\u5024\u884c\u5217\u56e0\u5b50\u5206\u89e3\u3092\u5fdc\u7528\u3057\u305f\u591a\u30af\u30e9\u30b9\u753b\u50cf\u5206\u985e\u30e2\u30c7\u30eb\u3092\u63d0\u6848\u3057, \u624b\u66f8\u304d\u6570\u5b57\u753b\u50cf\u3092\u5206\u985e\u3059\u308b\u3068\u3044\u3046\u554f\u984c\u306b\u5bfe\u3057\u3066\u63d0\u6848\u30e2\u30c7\u30eb\u3068\u5168\u7d50\u5408\u30cb\u30e5\u30fc\u30e9\u30eb\u30cd\u30c3\u30c8\u30ef\u30fc\u30af\u3067\u305d\u306e\u6027\u80fd\u3092\u6bd4\u8f03\u3059\u308b\u8ad6\u6587\u3092\u7d39\u4ecb\u3057\u307e\u3057\u305f\u3002\u672c\u8a18\u4e8b\u3067\u306f\u3001\u305d\u306e\u30a2\u30eb\u30b4\u30ea\u30ba\u30e0\u3092\u5b9f\u88c5\u3057\u3001\u5143\u8ad6\u6587\u306e\u518d\u73fe\u5b9f\u9a13\u3092\u884c\u3044\u307e\u3059\u3002","robots":{"index":"index","follow":"follow","max-snippet":"max-snippet:-1","max-image-preview":"max-image-preview:large","max-video-preview":"max-video-preview:-1"},"canonical":"https:\/\/qard.is.tohoku.ac.jp\/T-Wave\/2026\/02\/13\/\u3010\u5b9f\u8df5\u7de8\u3011\u91cf\u5b50\u30a2\u30cb\u30fc\u30ea\u30f3\u30b0\u3092\u6d3b\u7528\u3057\u305f\u975e\u8ca0\u4e8c\u5024\/","og_locale":"ja_JP","og_type":"article","og_title":"\u3010\u5b9f\u8df5\u7de8\u3011\u91cf\u5b50\u30a2\u30cb\u30fc\u30ea\u30f3\u30b0\u3092\u6d3b\u7528\u3057\u305f\u975e\u8ca0\u4e8c\u5024\u884c\u5217\u56e0\u5b50\u5206\u89e3\u306b\u3088\u308b\u753b\u50cf\u5206\u985e - T-QARD Harbor","og_description":"\u89e3\u8aac\u8a18\u4e8b\u300c\u91cf\u5b50\u30a2\u30cb\u30fc\u30ea\u30f3\u30b0\u3092\u6d3b\u7528\u3057\u305f\u975e\u8ca0\u4e8c\u5024\u884c\u5217\u56e0\u5b50\u5206\u89e3\u306b\u3088\u308b\u753b\u50cf\u5206\u985e\u300d\u3067\u306f\u3001\u975e\u8ca0\u4e8c\u5024\u884c\u5217\u56e0\u5b50\u5206\u89e3\u3092\u5fdc\u7528\u3057\u305f\u591a\u30af\u30e9\u30b9\u753b\u50cf\u5206\u985e\u30e2\u30c7\u30eb\u3092\u63d0\u6848\u3057, \u624b\u66f8\u304d\u6570\u5b57\u753b\u50cf\u3092\u5206\u985e\u3059\u308b\u3068\u3044\u3046\u554f\u984c\u306b\u5bfe\u3057\u3066\u63d0\u6848\u30e2\u30c7\u30eb\u3068\u5168\u7d50\u5408\u30cb\u30e5\u30fc\u30e9\u30eb\u30cd\u30c3\u30c8\u30ef\u30fc\u30af\u3067\u305d\u306e\u6027\u80fd\u3092\u6bd4\u8f03\u3059\u308b\u8ad6\u6587\u3092\u7d39\u4ecb\u3057\u307e\u3057\u305f\u3002\u672c\u8a18\u4e8b\u3067\u306f\u3001\u305d\u306e\u30a2\u30eb\u30b4\u30ea\u30ba\u30e0\u3092\u5b9f\u88c5\u3057\u3001\u5143\u8ad6\u6587\u306e\u518d\u73fe\u5b9f\u9a13\u3092\u884c\u3044\u307e\u3059\u3002","og_url":"https:\/\/qard.is.tohoku.ac.jp\/T-Wave\/2026\/02\/13\/\u3010\u5b9f\u8df5\u7de8\u3011\u91cf\u5b50\u30a2\u30cb\u30fc\u30ea\u30f3\u30b0\u3092\u6d3b\u7528\u3057\u305f\u975e\u8ca0\u4e8c\u5024\/","og_site_name":"T-QARD Harbor","article_published_time":"2026-02-13T01:12:46+00:00","og_image":[{"width":1500,"height":1213,"url":"https:\/\/qard.is.tohoku.ac.jp\/T-Wave\/wp-content\/uploads\/2026\/02\/b7cd83984219402b84f667ac12783404.png","type":"image\/png"}],"author":"Ryuga Yamada","twitter_card":"summary_large_image","twitter_misc":{"\u57f7\u7b46\u8005":"Ryuga Yamada","\u63a8\u5b9a\u8aad\u307f\u53d6\u308a\u6642\u9593":"161\u5206"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"Article","@id":"https:\/\/qard.is.tohoku.ac.jp\/T-Wave\/2026\/02\/13\/%e3%80%90%e5%ae%9f%e8%b7%b5%e7%b7%a8%e3%80%91%e9%87%8f%e5%ad%90%e3%82%a2%e3%83%8b%e3%83%bc%e3%83%aa%e3%83%b3%e3%82%b0%e3%82%92%e6%b4%bb%e7%94%a8%e3%81%97%e3%81%9f%e9%9d%9e%e8%b2%a0%e4%ba%8c%e5%80%a4\/#article","isPartOf":{"@id":"https:\/\/qard.is.tohoku.ac.jp\/T-Wave\/2026\/02\/13\/%e3%80%90%e5%ae%9f%e8%b7%b5%e7%b7%a8%e3%80%91%e9%87%8f%e5%ad%90%e3%82%a2%e3%83%8b%e3%83%bc%e3%83%aa%e3%83%b3%e3%82%b0%e3%82%92%e6%b4%bb%e7%94%a8%e3%81%97%e3%81%9f%e9%9d%9e%e8%b2%a0%e4%ba%8c%e5%80%a4\/"},"author":{"name":"Ryuga Yamada","@id":"https:\/\/qard.is.tohoku.ac.jp\/T-Wave\/#\/schema\/person\/be616b43e2d5a6ad431f358bc5bd34f4"},"headline":"\u3010\u5b9f\u8df5\u7de8\u3011\u91cf\u5b50\u30a2\u30cb\u30fc\u30ea\u30f3\u30b0\u3092\u6d3b\u7528\u3057\u305f\u975e\u8ca0\u4e8c\u5024\u884c\u5217\u56e0\u5b50\u5206\u89e3\u306b\u3088\u308b\u753b\u50cf\u5206\u985e","datePublished":"2026-02-13T01:12:46+00:00","mainEntityOfPage":{"@id":"https:\/\/qard.is.tohoku.ac.jp\/T-Wave\/2026\/02\/13\/%e3%80%90%e5%ae%9f%e8%b7%b5%e7%b7%a8%e3%80%91%e9%87%8f%e5%ad%90%e3%82%a2%e3%83%8b%e3%83%bc%e3%83%aa%e3%83%b3%e3%82%b0%e3%82%92%e6%b4%bb%e7%94%a8%e3%81%97%e3%81%9f%e9%9d%9e%e8%b2%a0%e4%ba%8c%e5%80%a4\/"},"wordCount":383,"commentCount":0,"image":{"@id":"https:\/\/qard.is.tohoku.ac.jp\/T-Wave\/2026\/02\/13\/%e3%80%90%e5%ae%9f%e8%b7%b5%e7%b7%a8%e3%80%91%e9%87%8f%e5%ad%90%e3%82%a2%e3%83%8b%e3%83%bc%e3%83%aa%e3%83%b3%e3%82%b0%e3%82%92%e6%b4%bb%e7%94%a8%e3%81%97%e3%81%9f%e9%9d%9e%e8%b2%a0%e4%ba%8c%e5%80%a4\/#primaryimage"},"thumbnailUrl":"https:\/\/qard.is.tohoku.ac.jp\/T-Wave\/wp-content\/uploads\/2026\/02\/b7cd83984219402b84f667ac12783404.png","keywords":["\u30cb\u30e5\u30fc\u30e9\u30eb\u30cd\u30c3\u30c8\u30ef\u30fc\u30af","\u5b9f\u8df5\u8a18\u4e8b","\u753b\u50cf\u5206\u985e","\u91cf\u5b50\u30a2\u30cb\u30fc\u30ea\u30f3\u30b0","\u975e\u8ca0\u4e8c\u5024\u884c\u5217\u56e0\u5b50\u5206\u89e3"],"articleSection":["\u5b9f\u8df5\u8a18\u4e8b"],"inLanguage":"ja","potentialAction":[{"@type":"CommentAction","name":"Comment","target":["https:\/\/qard.is.tohoku.ac.jp\/T-Wave\/2026\/02\/13\/%e3%80%90%e5%ae%9f%e8%b7%b5%e7%b7%a8%e3%80%91%e9%87%8f%e5%ad%90%e3%82%a2%e3%83%8b%e3%83%bc%e3%83%aa%e3%83%b3%e3%82%b0%e3%82%92%e6%b4%bb%e7%94%a8%e3%81%97%e3%81%9f%e9%9d%9e%e8%b2%a0%e4%ba%8c%e5%80%a4\/#respond"]}]},{"@type":"WebPage","@id":"https:\/\/qard.is.tohoku.ac.jp\/T-Wave\/2026\/02\/13\/%e3%80%90%e5%ae%9f%e8%b7%b5%e7%b7%a8%e3%80%91%e9%87%8f%e5%ad%90%e3%82%a2%e3%83%8b%e3%83%bc%e3%83%aa%e3%83%b3%e3%82%b0%e3%82%92%e6%b4%bb%e7%94%a8%e3%81%97%e3%81%9f%e9%9d%9e%e8%b2%a0%e4%ba%8c%e5%80%a4\/","url":"https:\/\/qard.is.tohoku.ac.jp\/T-Wave\/2026\/02\/13\/%e3%80%90%e5%ae%9f%e8%b7%b5%e7%b7%a8%e3%80%91%e9%87%8f%e5%ad%90%e3%82%a2%e3%83%8b%e3%83%bc%e3%83%aa%e3%83%b3%e3%82%b0%e3%82%92%e6%b4%bb%e7%94%a8%e3%81%97%e3%81%9f%e9%9d%9e%e8%b2%a0%e4%ba%8c%e5%80%a4\/","name":"\u3010\u5b9f\u8df5\u7de8\u3011\u91cf\u5b50\u30a2\u30cb\u30fc\u30ea\u30f3\u30b0\u3092\u6d3b\u7528\u3057\u305f\u975e\u8ca0\u4e8c\u5024\u884c\u5217\u56e0\u5b50\u5206\u89e3\u306b\u3088\u308b\u753b\u50cf\u5206\u985e - T-QARD Harbor","isPartOf":{"@id":"https:\/\/qard.is.tohoku.ac.jp\/T-Wave\/#website"},"primaryImageOfPage":{"@id":"https:\/\/qard.is.tohoku.ac.jp\/T-Wave\/2026\/02\/13\/%e3%80%90%e5%ae%9f%e8%b7%b5%e7%b7%a8%e3%80%91%e9%87%8f%e5%ad%90%e3%82%a2%e3%83%8b%e3%83%bc%e3%83%aa%e3%83%b3%e3%82%b0%e3%82%92%e6%b4%bb%e7%94%a8%e3%81%97%e3%81%9f%e9%9d%9e%e8%b2%a0%e4%ba%8c%e5%80%a4\/#primaryimage"},"image":{"@id":"https:\/\/qard.is.tohoku.ac.jp\/T-Wave\/2026\/02\/13\/%e3%80%90%e5%ae%9f%e8%b7%b5%e7%b7%a8%e3%80%91%e9%87%8f%e5%ad%90%e3%82%a2%e3%83%8b%e3%83%bc%e3%83%aa%e3%83%b3%e3%82%b0%e3%82%92%e6%b4%bb%e7%94%a8%e3%81%97%e3%81%9f%e9%9d%9e%e8%b2%a0%e4%ba%8c%e5%80%a4\/#primaryimage"},"thumbnailUrl":"https:\/\/qard.is.tohoku.ac.jp\/T-Wave\/wp-content\/uploads\/2026\/02\/b7cd83984219402b84f667ac12783404.png","datePublished":"2026-02-13T01:12:46+00:00","author":{"@id":"https:\/\/qard.is.tohoku.ac.jp\/T-Wave\/#\/schema\/person\/be616b43e2d5a6ad431f358bc5bd34f4"},"description":"\u89e3\u8aac\u8a18\u4e8b\u300c\u91cf\u5b50\u30a2\u30cb\u30fc\u30ea\u30f3\u30b0\u3092\u6d3b\u7528\u3057\u305f\u975e\u8ca0\u4e8c\u5024\u884c\u5217\u56e0\u5b50\u5206\u89e3\u306b\u3088\u308b\u753b\u50cf\u5206\u985e\u300d\u3067\u306f\u3001\u975e\u8ca0\u4e8c\u5024\u884c\u5217\u56e0\u5b50\u5206\u89e3\u3092\u5fdc\u7528\u3057\u305f\u591a\u30af\u30e9\u30b9\u753b\u50cf\u5206\u985e\u30e2\u30c7\u30eb\u3092\u63d0\u6848\u3057, \u624b\u66f8\u304d\u6570\u5b57\u753b\u50cf\u3092\u5206\u985e\u3059\u308b\u3068\u3044\u3046\u554f\u984c\u306b\u5bfe\u3057\u3066\u63d0\u6848\u30e2\u30c7\u30eb\u3068\u5168\u7d50\u5408\u30cb\u30e5\u30fc\u30e9\u30eb\u30cd\u30c3\u30c8\u30ef\u30fc\u30af\u3067\u305d\u306e\u6027\u80fd\u3092\u6bd4\u8f03\u3059\u308b\u8ad6\u6587\u3092\u7d39\u4ecb\u3057\u307e\u3057\u305f\u3002\u672c\u8a18\u4e8b\u3067\u306f\u3001\u305d\u306e\u30a2\u30eb\u30b4\u30ea\u30ba\u30e0\u3092\u5b9f\u88c5\u3057\u3001\u5143\u8ad6\u6587\u306e\u518d\u73fe\u5b9f\u9a13\u3092\u884c\u3044\u307e\u3059\u3002","breadcrumb":{"@id":"https:\/\/qard.is.tohoku.ac.jp\/T-Wave\/2026\/02\/13\/%e3%80%90%e5%ae%9f%e8%b7%b5%e7%b7%a8%e3%80%91%e9%87%8f%e5%ad%90%e3%82%a2%e3%83%8b%e3%83%bc%e3%83%aa%e3%83%b3%e3%82%b0%e3%82%92%e6%b4%bb%e7%94%a8%e3%81%97%e3%81%9f%e9%9d%9e%e8%b2%a0%e4%ba%8c%e5%80%a4\/#breadcrumb"},"inLanguage":"ja","potentialAction":[{"@type":"ReadAction","target":["https:\/\/qard.is.tohoku.ac.jp\/T-Wave\/2026\/02\/13\/%e3%80%90%e5%ae%9f%e8%b7%b5%e7%b7%a8%e3%80%91%e9%87%8f%e5%ad%90%e3%82%a2%e3%83%8b%e3%83%bc%e3%83%aa%e3%83%b3%e3%82%b0%e3%82%92%e6%b4%bb%e7%94%a8%e3%81%97%e3%81%9f%e9%9d%9e%e8%b2%a0%e4%ba%8c%e5%80%a4\/"]}]},{"@type":"ImageObject","inLanguage":"ja","@id":"https:\/\/qard.is.tohoku.ac.jp\/T-Wave\/2026\/02\/13\/%e3%80%90%e5%ae%9f%e8%b7%b5%e7%b7%a8%e3%80%91%e9%87%8f%e5%ad%90%e3%82%a2%e3%83%8b%e3%83%bc%e3%83%aa%e3%83%b3%e3%82%b0%e3%82%92%e6%b4%bb%e7%94%a8%e3%81%97%e3%81%9f%e9%9d%9e%e8%b2%a0%e4%ba%8c%e5%80%a4\/#primaryimage","url":"https:\/\/qard.is.tohoku.ac.jp\/T-Wave\/wp-content\/uploads\/2026\/02\/b7cd83984219402b84f667ac12783404.png","contentUrl":"https:\/\/qard.is.tohoku.ac.jp\/T-Wave\/wp-content\/uploads\/2026\/02\/b7cd83984219402b84f667ac12783404.png","width":1500,"height":1213},{"@type":"BreadcrumbList","@id":"https:\/\/qard.is.tohoku.ac.jp\/T-Wave\/2026\/02\/13\/%e3%80%90%e5%ae%9f%e8%b7%b5%e7%b7%a8%e3%80%91%e9%87%8f%e5%ad%90%e3%82%a2%e3%83%8b%e3%83%bc%e3%83%aa%e3%83%b3%e3%82%b0%e3%82%92%e6%b4%bb%e7%94%a8%e3%81%97%e3%81%9f%e9%9d%9e%e8%b2%a0%e4%ba%8c%e5%80%a4\/#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"\u30db\u30fc\u30e0","item":"https:\/\/qard.is.tohoku.ac.jp\/T-Wave\/"},{"@type":"ListItem","position":2,"name":"\u3010\u5b9f\u8df5\u7de8\u3011\u91cf\u5b50\u30a2\u30cb\u30fc\u30ea\u30f3\u30b0\u3092\u6d3b\u7528\u3057\u305f\u975e\u8ca0\u4e8c\u5024\u884c\u5217\u56e0\u5b50\u5206\u89e3\u306b\u3088\u308b\u753b\u50cf\u5206\u985e"}]},{"@type":"WebSite","@id":"https:\/\/qard.is.tohoku.ac.jp\/T-Wave\/#website","url":"https:\/\/qard.is.tohoku.ac.jp\/T-Wave\/","name":"T-QARD Harbor","description":"T-QARD Harbor\u306f\u6771\u5317\u5927\u5b66\u91cf\u5b50\u30a2\u30d7\u30ea\u30b1\u30fc\u30b7\u30e7\u30f3\u7814\u7a76\u958b\u767a\u30bb\u30f3\u30bf\u30fc\u5b66\u751f\u30c1\u30fc\u30e0\u300cT-QARD Crews\u300d\u304c\u904b\u55b6\u3059\u308b\u3001 \u6570\u7406\u60c5\u5831\u7d71\u8a08\u3001\u91cf\u5b50\u60c5\u5831\u3001\u6700\u9069\u5316\u3001\u6a5f\u68b0\u5b66\u7fd2\u5206\u91ce\u306e\u60c5\u5831\u3092\u63d0\u4f9b\u3059\u308bWeb\u30b5\u30a4\u30c8\u3067\u3059","potentialAction":[{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https:\/\/qard.is.tohoku.ac.jp\/T-Wave\/?s={search_term_string}"},"query-input":{"@type":"PropertyValueSpecification","valueRequired":true,"valueName":"search_term_string"}}],"inLanguage":"ja"},{"@type":"Person","@id":"https:\/\/qard.is.tohoku.ac.jp\/T-Wave\/#\/schema\/person\/be616b43e2d5a6ad431f358bc5bd34f4","name":"Ryuga Yamada","image":{"@type":"ImageObject","inLanguage":"ja","@id":"https:\/\/secure.gravatar.com\/avatar\/6f1965735667140ac51f351536e86b546a2ca3406c60bbf23e27ae7de1091db1?s=96&d=mm&r=g","url":"https:\/\/secure.gravatar.com\/avatar\/6f1965735667140ac51f351536e86b546a2ca3406c60bbf23e27ae7de1091db1?s=96&d=mm&r=g","contentUrl":"https:\/\/secure.gravatar.com\/avatar\/6f1965735667140ac51f351536e86b546a2ca3406c60bbf23e27ae7de1091db1?s=96&d=mm&r=g","caption":"Ryuga Yamada"},"description":"\u6771\u5317\u5927\u5b66 \u5de5\u5b66\u90e8 \u5b66\u58eb\u8ab2\u7a0b","url":"https:\/\/qard.is.tohoku.ac.jp\/T-Wave\/author\/ryuga-yamada\/"}]}},"_links":{"self":[{"href":"https:\/\/qard.is.tohoku.ac.jp\/T-Wave\/wp-json\/wp\/v2\/posts\/8643","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/qard.is.tohoku.ac.jp\/T-Wave\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/qard.is.tohoku.ac.jp\/T-Wave\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/qard.is.tohoku.ac.jp\/T-Wave\/wp-json\/wp\/v2\/users\/20"}],"replies":[{"embeddable":true,"href":"https:\/\/qard.is.tohoku.ac.jp\/T-Wave\/wp-json\/wp\/v2\/comments?post=8643"}],"version-history":[{"count":11,"href":"https:\/\/qard.is.tohoku.ac.jp\/T-Wave\/wp-json\/wp\/v2\/posts\/8643\/revisions"}],"predecessor-version":[{"id":8659,"href":"https:\/\/qard.is.tohoku.ac.jp\/T-Wave\/wp-json\/wp\/v2\/posts\/8643\/revisions\/8659"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/qard.is.tohoku.ac.jp\/T-Wave\/wp-json\/wp\/v2\/media\/8650"}],"wp:attachment":[{"href":"https:\/\/qard.is.tohoku.ac.jp\/T-Wave\/wp-json\/wp\/v2\/media?parent=8643"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/qard.is.tohoku.ac.jp\/T-Wave\/wp-json\/wp\/v2\/categories?post=8643"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/qard.is.tohoku.ac.jp\/T-Wave\/wp-json\/wp\/v2\/tags?post=8643"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}