{"id":932,"date":"2024-02-15T05:13:15","date_gmt":"2024-02-14T20:13:15","guid":{"rendered":"https:\/\/www.dogrow.net\/nnet\/?p=932"},"modified":"2025-03-13T08:59:05","modified_gmt":"2025-03-12T23:59:05","slug":"blog29","status":"publish","type":"post","link":"https:\/\/www.dogrow.net\/nnet\/blog29\/","title":{"rendered":"(29) K-fold cross-validation\u7528\u306e\u30c7\u30fc\u30bf\u30bb\u30c3\u30c8\u3092\u4f5c\u308b\u3002"},"content":{"rendered":"<h1 class=\"my_h\">\u30101\u3011\u3084\u308a\u305f\u3044\u3053\u3068<\/h1>\n<p>\u6a5f\u68b0\u5b66\u7fd2\u306b\u4f7f\u3048\u308b\u30c7\u30fc\u30bf\u304c 5,000\u4ef6\u3042\u308b\u3002<br \/>\n<span class=\"my_fc_deeppinkBBig\">K-fold cross-validation<\/span>\uff08\uff1dK\u5206\u5272\u4ea4\u5dee\u691c\u8a3c\uff09<br \/>\n\u3092\u884c\u3046\u305f\u3081\u3001\u3053\u306e\u30c7\u30fc\u30bf\u3092 5\u30bb\u30c3\u30c8\u306b\u5206\u3051\u305f\u3044\u3002<\/p>\n<p>\u6761\u4ef6\u306f\u4ee5\u4e0b\u306e\u901a\u308a\u3002<br \/>\n\u30fb\u5404\u56de\u306e\u8a13\u7df4\u306e\u4e2d\u3067\u3001\u8a13\u7df4\u7528\u30c7\u30fc\u30bf\u3068\u30c6\u30b9\u30c8\u7528\u30c7\u30fc\u30bf\u304c\u91cd\u8907\u3057\u3066\u306f\u306a\u3089\u306a\u3044\u3002<br \/>\n\u30fb\u5404\u56de\u306e\u8a13\u7df4\u306e\u9593\u3067\u3001\u30c6\u30b9\u30c8\u7528\u30c7\u30fc\u30bf\u304c\u91cd\u8907\u3057\u3066\u306f\u306a\u3089\u306a\u3044\u3002<\/p>\n<p>\u3053\u306e\u6761\u4ef6\u3092\u6e80\u305f\u3059\u30c7\u30fc\u30bf\u3092\u3001Python\u3092\u4f7f\u3063\u3066\u30b5\u30af\u30c3\u3068\u4f5c\u308a\u305f\u3044\u3002<\/p>\n<h1 class=\"my_h\">\u30102\u3011\u3084\u3063\u3066\u307f\u308b<\/h1>\n<h2 class=\"my_h\">1) Python shell\u3067\u4e00\u884c\u305a\u3064\u5b9f\u884c\u3057\u3066\u307f\u308b\u3002<\/h2>\n<p>\u307e\u305a\u306f\u30015,000\u4ef6\u306e\u30c7\u30fc\u30bf\u304c\u683c\u7d0d\u3055\u308c\u3066\u3044\u308b\u30d5\u30a1\u30a4\u30eb\u3092\u30ed\u30fc\u30c9\u3059\u308b\u3002<\/p>\n<pre class=\"brush: python; title: ; notranslate\" title=\"\">\r\nimport pandas as pd\r\nimport random\r\n\r\ndataAll = pd.read_csv('dataAll.csv', header=None)\r\n<\/pre>\n<p>\u6b21\u306b\u30011\u56de\u306e\u8a13\u7df4\u3067\u4f7f\u7528\u3059\u308b\u8a13\u7df4\u7528\uff0b\u691c\u8a3c\u7528\u30c7\u30fc\u30bf\u30bb\u30c3\u30c8\u3092\u4f5c\u308b\u3002<\/p>\n<pre class=\"brush: python; title: ; notranslate\" title=\"\">\r\nnumTest = 2\r\ni = 1\r\n\r\ndataNotUsed = dataAll               # \u672a\u4f7f\u7528\u30c7\u30fc\u30bf\u7fa4 (\u30c6\u30b9\u30c8\u7528\u306b\u672a\u4f7f\u7528)\r\ndataUsed    = pd.DataFrame()        # \u4f7f\u7528\u6e08\u30c7\u30fc\u30bf\u7fa4 (\u30c6\u30b9\u30c8\u7528\u306b\u4f7f\u7528\u6e08)\r\n\r\n# \u30c6\u30b9\u30c8\u7528\u30c7\u30fc\u30bf\u3092\u3001\u672a\u4f7f\u7528\u30c7\u30fc\u30bf\u7fa4\u304b\u3089\u53d6\u5f97\r\nnumNotUsed = len(dataNotUsed)       \r\niTest      = random.sample(range(numNotUsed), numTest)\r\ndataTest   = dataNotUsed.iloc&#x5B;iTest]\r\n\r\n# iTest\u3067\u53d6\u5f97\u3057\u305f\u884c\u306e\u30a4\u30f3\u30c7\u30c3\u30af\u30b9\u3092\u53d6\u5f97\r\nindex_to_drop = dataNotUsed.iloc&#x5B;iTest].index\r\n# \u3053\u308c\u3089\u306e\u30a4\u30f3\u30c7\u30c3\u30af\u30b9\u3092\u4f7f\u3063\u3066\u3001\u672a\u4f7f\u7528\u30c7\u30fc\u30bf\u7fa4\u304b\u3089\u4eca\u56de\u4f7f\u7528\u3059\u308b\u30c6\u30b9\u30c8\u7528\u30c7\u30fc\u30bf\u3092\u524a\u9664\r\ndataNotUsed   = dataNotUsed.drop(index_to_drop)\r\n\r\n# \u8a13\u7df4\u7528\u30c7\u30fc\u30bf\u3092\u4f5c\u6210\uff08\u672a\u4f7f\u7528\u30c7\u30fc\u30bf\uff0b\u4f7f\u7528\u6e08\u30c7\u30fc\u30bf\uff09\r\n# \u203b\u521d\u56de\u306a\u306e\u3067 dataUsed \u306f\u7a7a\u3063\u307d\r\ndataTrain = pd.concat(&#x5B;dataNotUsed, dataUsed], axis=0)\r\n\r\n# \u4f7f\u7528\u6e08\u30c7\u30fc\u30bf\u7fa4\u3092\u66f4\u65b0\uff08\u4f7f\u7528\u6e08\u307f\u30c7\u30fc\u30bf\uff0b\u4eca\u56de\u306e\u30c6\u30b9\u30c8\u7528\u30c7\u30fc\u30bf\uff09\r\ndataUsed  = pd.concat(&#x5B;dataUsed, dataTest], axis=0)\r\n\r\n# \u30d5\u30a1\u30a4\u30eb\u51fa\u529b \u203b\u76ee\u8996\u3067\u78ba\u8a8d\u3057\u3084\u3059\u3044\u3088\u3046\u306b 1\u5217\u76ee\u306e\u5024\u3067\u6607\u9806\u306b\u30bd\u30fc\u30c8\r\ndataTrain = dataTrain.sort_values(by=0, ascending=True)\r\ndataTest  = dataTest. sort_values(by=0, ascending=True)\r\ndataTrain.to_csv(f&quot;dataTrain_{i}.csv&quot;, header=None)\r\ndataTest. to_csv(f&quot;dataTest_{i}.csv&quot;,  header=None)\r\n\r\ni += 1\r\n<\/pre>\n<p><a href=\"https:\/\/chatgpt.com\/share\/67d21a37-2a1c-8007-9b8d-24acc084de8e\" target=\"_blank\">\u8a73\u7d30\u306f\u3001\u3053\u3061\u3089\u306e ChatGPT\u5927\u5148\u751f\u306e\u89e3\u8aac\u3092\u3054\u53c2\u7167\u304f\u3060\u3055\u3044\u3002<\/a><\/p>\n<h2 class=\"my_h\">2) \u95a2\u6570\u5316\u3057\u3001\u4e00\u6c17\u306b 5\u30bb\u30c3\u30c8\u3092\u751f\u6210\u3059\u308b\u3002<\/h2>\n<pre class=\"brush: python; title: ; notranslate\" title=\"\">\r\nimport pandas as pd\r\nimport random\r\n\r\ndef makeDataSet( dataAll_csv, numKFold, numTest ):\r\n    # \u6307\u5b9a CSV\u30d5\u30a1\u30a4\u30eb\u3092\u8aad\u307f\u8fbc\u307f\r\n    dataAll = pd.read_csv(dataAll_csv, header=None)\r\n    \r\n    # \u5909\u6570\u521d\u671f\u5316\r\n    dataNotUsed = dataAll               # \u672a\u4f7f\u7528\u30c7\u30fc\u30bf\u7fa4 (\u30c6\u30b9\u30c8\u7528\u306b\u672a\u4f7f\u7528)\r\n    dataUsed    = pd.DataFrame()        # \u4f7f\u7528\u6e08\u30c7\u30fc\u30bf\u7fa4 (\u30c6\u30b9\u30c8\u7528\u306b\u4f7f\u7528\u6e08)\r\n    \r\n    for i in range(numKFold):\r\n        # \u30c6\u30b9\u30c8\u7528\u30c7\u30fc\u30bf\u3092\u3001\u672a\u4f7f\u7528\u30c7\u30fc\u30bf\u7fa4\u304b\u3089\u53d6\u5f97\r\n        numNotUsed = len(dataNotUsed)\r\n        iTest    = random.sample(range(numNotUsed), numTest)\r\n        dataTest = dataNotUsed.iloc&#x5B;iTest]\r\n        \r\n        # iTest\u3067\u53d6\u5f97\u3057\u305f\u884c\u306e\u30a4\u30f3\u30c7\u30c3\u30af\u30b9\u3092\u53d6\u5f97\r\n        index_to_drop = dataNotUsed.iloc&#x5B;iTest].index\r\n        # \u3053\u308c\u3089\u306e\u30a4\u30f3\u30c7\u30c3\u30af\u30b9\u3092\u4f7f\u3063\u3066\u3001\u672a\u4f7f\u7528\u30c7\u30fc\u30bf\u7fa4\u304b\u3089\u4eca\u56de\u4f7f\u7528\u3059\u308b\u30c6\u30b9\u30c8\u7528\u30c7\u30fc\u30bf\u3092\u524a\u9664\r\n        dataNotUsed   = dataNotUsed.drop(index_to_drop)\r\n        \r\n        # \u8a13\u7df4\u7528\u30c7\u30fc\u30bf\u3092\u4f5c\u6210\uff08\u672a\u4f7f\u7528\u30c7\u30fc\u30bf\uff0b\u4f7f\u7528\u6e08\u30c7\u30fc\u30bf\uff09\r\n        dataTrain = pd.concat(&#x5B;dataNotUsed, dataUsed], axis=0)\r\n        \r\n        # \u4f7f\u7528\u6e08\u30c7\u30fc\u30bf\u7fa4\u3092\u66f4\u65b0\uff08\u4f7f\u7528\u6e08\u307f\u30c7\u30fc\u30bf\uff0b\u4eca\u56de\u306e\u30c6\u30b9\u30c8\u7528\u30c7\u30fc\u30bf\uff09\r\n        dataUsed  = pd.concat(&#x5B;dataUsed, dataTest], axis=0)\r\n        \r\n        # \u30d5\u30a1\u30a4\u30eb\u51fa\u529b \u203b\u76ee\u8996\u3067\u78ba\u8a8d\u3057\u3084\u3059\u3044\u3088\u3046\u306b 1\u5217\u76ee\u306e\u5024\u3067\u6607\u9806\u306b\u30bd\u30fc\u30c8\r\n        dataTrain = dataTrain.sort_values(by=0, ascending=True)\r\n        dataTest  = dataTest. sort_values(by=0, ascending=True)\r\n        dataTrain.to_csv(f&quot;dataTrain_{i}.csv&quot;, header=None)\r\n        dataTest. to_csv(f&quot;dataTest_{i}.csv&quot;,  header=None)\r\n\r\n# \u30c7\u30fc\u30bf\u30d5\u30a1\u30a4\u30eb\u540d\u3092\u6307\u5b9a\u3057\u3066\u6a5f\u80fd\u3092\u5b9f\u884c\r\nmakeDataSet('dataAll.csv', 5, 500)\r\n<\/pre>\n<hr class=\"my_hr_bottom\">\n","protected":false},"excerpt":{"rendered":"<p>\u30101\u3011\u3084\u308a\u305f\u3044\u3053\u3068 \u6a5f\u68b0\u5b66\u7fd2\u306b\u4f7f\u3048\u308b\u30c7\u30fc\u30bf\u304c 5,000\u4ef6\u3042\u308b\u3002 K-fold cross-validation\uff08\uff1dK\u5206\u5272\u4ea4\u5dee\u691c\u8a3c\uff09 \u3092\u884c\u3046\u305f\u3081\u3001\u3053\u306e\u30c7\u30fc\u30bf\u3092 5\u30bb\u30c3\u30c8\u306b\u5206\u3051\u305f\u3044\u3002 \u6761\u4ef6\u306f\u4ee5\u4e0b\u306e\u901a\u308a\u3002 \u30fb\u5404\u56de\u306e\u8a13\u7df4\u306e\u4e2d\u3067\u2026 <span class=\"read-more\"><a href=\"https:\/\/www.dogrow.net\/nnet\/blog29\/\">\u7d9a\u304d\u3092\u8aad\u3080 &raquo;<\/a><\/span><\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[2],"tags":[],"class_list":["post-932","post","type-post","status-publish","format-standard","hentry","category-2"],"views":855,"amp_enabled":true,"_links":{"self":[{"href":"https:\/\/www.dogrow.net\/nnet\/wp-json\/wp\/v2\/posts\/932","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.dogrow.net\/nnet\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.dogrow.net\/nnet\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.dogrow.net\/nnet\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/www.dogrow.net\/nnet\/wp-json\/wp\/v2\/comments?post=932"}],"version-history":[{"count":9,"href":"https:\/\/www.dogrow.net\/nnet\/wp-json\/wp\/v2\/posts\/932\/revisions"}],"predecessor-version":[{"id":945,"href":"https:\/\/www.dogrow.net\/nnet\/wp-json\/wp\/v2\/posts\/932\/revisions\/945"}],"wp:attachment":[{"href":"https:\/\/www.dogrow.net\/nnet\/wp-json\/wp\/v2\/media?parent=932"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.dogrow.net\/nnet\/wp-json\/wp\/v2\/categories?post=932"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.dogrow.net\/nnet\/wp-json\/wp\/v2\/tags?post=932"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}