{"id":19020,"date":"2026-09-03T08:34:27","date_gmt":"2026-09-02T22:34:27","guid":{"rendered":"https:\/\/geoffreychen.com\/regularisation-trades-perfect-fit-for-reliable-generalisation\/"},"modified":"2026-09-03T08:34:27","modified_gmt":"2026-09-02T22:34:27","slug":"regularisation-trades-perfect-fit-for-reliable-generalisation","status":"publish","type":"post","link":"https:\/\/geoffreychen.com\/zh\/regularisation-trades-perfect-fit-for-reliable-generalisation\/","title":{"rendered":"\u6b63\u5219\u5316\uff1a\u4e3a\u4ec0\u4e48\u5c11\u8ffd\u6c42\u4e00\u70b9\u62df\u5408\u53cd\u800c\u66f4\u53ef\u9760\uff1f"},"content":{"rendered":"<\/p>\n

\"\u6b63\u5219\u5316\uff1a\u4e3a\u4ec0\u4e48\u5c11\u8ffd\u6c42\u4e00\u70b9\u62df\u5408\u53cd\u800c\u66f4\u53ef\u9760\uff1f\"<\/p>\n

\u6b63\u5219\u5316\u662f\u5728\u8bad\u7ec3\u6a21\u578b\u65f6\uff0c\u9664\u4e86\u8981\u6c42\u5b83\u89e3\u91ca\u5df2\u6709\u6570\u636e\uff0c\u8fd8\u5bf9\u8fc7\u5ea6\u590d\u6742\u7684\u89e3\u6cd5\u65bd\u52a0\u4ee3\u4ef7\u3002\u5b83\u4e3b\u52a8\u653e\u5f03\u5bf9\u8bad\u7ec3\u6837\u672c\u7684\u5b8c\u7f8e\u8fce\u5408\uff0c\u4ee5\u6362\u53d6\u9762\u5bf9\u65b0\u6570\u636e\u65f6\u66f4\u7a33\u5b9a\u7684\u5224\u65ad\u3002<\/p>\n

\u4f8b\u5982\uff0c\u7528\u4e00\u6761\u66f2\u7ebf\u7a7f\u8fc7\u5e26\u6709\u6d4b\u91cf\u8bef\u5dee\u7684\u6240\u6709\u6570\u636e\u70b9\uff0c\u8bad\u7ec3\u8bef\u5dee\u53ef\u4ee5\u964d\u5f97\u5f88\u4f4e\uff0c\u4f46\u66f2\u7ebf\u7684\u5267\u70c8\u5f2f\u6298\u53ef\u80fd\u53ea\u662f\u5728\u8bb0\u5fc6\u566a\u58f0\u3002\u52a0\u5165\u6b63\u5219\u5316\u540e\uff0c\u6a21\u578b\u503e\u5411\u9009\u62e9\u8f83\u5e73\u7f13\u7684\u66f2\u7ebf\uff1a\u5b83\u5728\u65e7\u6570\u636e\u4e0a\u672a\u5fc5\u6700\u7cbe\u786e\uff0c\u5374\u53ef\u80fd\u66f4\u63a5\u8fd1\u4ea7\u751f\u6570\u636e\u7684\u7a33\u5b9a\u5173\u7cfb\u3002<\/p>\n

\u6b63\u5219\u5316\u4e0d\u540c\u4e8e\u5f52\u4e00\u5316\uff1b\u540e\u8005\u4e3b\u8981\u8c03\u6574\u53d8\u91cf\u7684\u5c3a\u5ea6\u3002\u5b83\u4e5f\u4e0d\u7b49\u4e8e\u7b80\u5355\u5730\u7f29\u5c0f\u6a21\u578b\uff1a\u9650\u5236\u53ef\u4ee5\u60e9\u7f5a\u8fc7\u5927\u7684\u53c2\u6570\u3001\u8fc7\u591a\u7684\u7279\u5f81\u6216\u8fc7\u6df1\u7684\u7ed3\u6784\u3002\u9650\u5236\u8fc7\u5f31\u4ecd\u4f1a\u8fc7\u62df\u5408\uff0c\u8fc7\u5f3a\u5219\u4f1a\u6b20\u62df\u5408\u3002\u4e2d\u5fc3\u95ee\u9898\u4e0d\u662f\u201c\u8d8a\u7b80\u5355\u8d8a\u597d\u201d\uff0c\u800c\u662f\u54ea\u4e9b\u590d\u6742\u6027\u5f97\u5230\u8bc1\u636e\u652f\u6301\u3002\u53ef\u9760\u7684\u6a21\u578b\u5e76\u975e\u8bb0\u4f4f\u4e00\u5207\uff0c\u800c\u662f\u62d2\u7edd\u628a\u5076\u7136\u7ec6\u8282\u5f53\u4f5c\u89c4\u5f8b\u3002<\/p>\n

https:\/\/scikit-learn.org\/stable\/auto_examples\/model_selection\/plot_train_error_vs_test_error.html<\/a><\/p>\n

","protected":false},"excerpt":{"rendered":"","protected":false},"author":17897162,"featured_media":0,"comment_status":"","ping_status":"","sticky":false,"template":"","format":"standard","meta":{"_coblocks_attr":"","_coblocks_dimensions":"","_coblocks_responsive_height":"","_coblocks_accordion_ie_support":"","_jetpack_newsletter_access":"","_jetpack_dont_email_post_to_subs":false,"_jetpack_newsletter_tier_id":0,"_jetpack_memberships_contains_paywalled_content":false,"_geoff_meta_description_en":"Regularisation limits model complexity, sacrificing a perfect training fit to resist noise and generalise more reliably, though too much can underfit.","_geoff_meta_description_zh":"\u6b63\u5219\u5316\u901a\u8fc7\u9650\u5236\u6a21\u578b\u590d\u6742\u5ea6\uff0c\u653e\u5f03\u5bf9\u8bad\u7ec3\u6837\u672c\u7684\u5b8c\u7f8e\u8fce\u5408\uff0c\u4ee5\u51cf\u5c11\u5076\u7136\u566a\u58f0\u5f71\u54cd\u5e76\u63d0\u9ad8\u5bf9\u672a\u89c1\u6570\u636e\u7684\u53ef\u9760\u6027\uff1b\u9650\u5236\u8fc7\u5f3a\u4e5f\u4f1a\u9020\u6210\u6b20\u62df\u5408\u3002","_wpcom_ai_launchpad_first_post":false,"_jetpack_feature_clip_id":0,"_jetpack_memberships_contains_paid_content":false,"footnotes":"","jetpack_publicize_message":"","jetpack_publicize_feature_enabled":true,"jetpack_social_post_already_shared":true,"jetpack_social_options":{"image_generator_settings":{"template":"highway","default_image_id":0,"font":"","enabled":false},"version":2},"jetpack_post_was_ever_published":false},"categories":[787294141],"tags":[787299126,787299124,787299127,787299125],"class_list":["post-19020","post","type-post","status-publish","format-standard","hentry","category-philosophical-concepts","tag-machine-learning","tag-regularisation","tag-machine-learning-zh","tag-regularisation-zh",""],"translatepress_titles":{"en_AU":"Regularisation: Why Can a Less Exact Fit Be More Reliable?","zh_CN":"\u6b63\u5219\u5316\uff1a\u4e3a\u4ec0\u4e48\u5c11\u8ffd\u6c42\u4e00\u70b9\u62df\u5408\u53cd\u800c\u66f4\u53ef\u9760\uff1f"},"jetpack_publicize_connections":[],"jetpack_likes_enabled":false,"jetpack_sharing_enabled":true,"jetpack_shortlink":"https:\/\/wp.me\/p1ar9H-4WM","jetpack_featured_media_url":"","_links":{"self":[{"href":"https:\/\/geoffreychen.com\/zh\/wp-json\/wp\/v2\/posts\/19020","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/geoffreychen.com\/zh\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/geoffreychen.com\/zh\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/geoffreychen.com\/zh\/wp-json\/wp\/v2\/users\/17897162"}],"replies":[{"embeddable":true,"href":"https:\/\/geoffreychen.com\/zh\/wp-json\/wp\/v2\/comments?post=19020"}],"version-history":[{"count":0,"href":"https:\/\/geoffreychen.com\/zh\/wp-json\/wp\/v2\/posts\/19020\/revisions"}],"wp:attachment":[{"href":"https:\/\/geoffreychen.com\/zh\/wp-json\/wp\/v2\/media?parent=19020"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/geoffreychen.com\/zh\/wp-json\/wp\/v2\/categories?post=19020"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/geoffreychen.com\/zh\/wp-json\/wp\/v2\/tags?post=19020"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}