{"id":490,"date":"2024-03-13T18:09:59","date_gmt":"2024-03-13T23:09:59","guid":{"rendered":"https:\/\/ozer.gt\/log\/?p=490"},"modified":"2024-07-17T18:02:12","modified_gmt":"2024-07-17T23:02:12","slug":"when-do-neural-nets-outperform-boosted-trees-on-tabular-data","status":"publish","type":"post","link":"https:\/\/ozer.gt\/log\/2024\/03\/13\/when-do-neural-nets-outperform-boosted-trees-on-tabular-data\/","title":{"rendered":"When do neural nets outperform boosted trees on tabular data?"},"content":{"rendered":"<p>Otherwise, tree ensembles continue to outperform neural networks. The decision tree in the figure shows the winner among the top five methods.<\/p>\n<p>Now, the background:<\/p>\n<p>I explored the why of this question before, but didn&#8217;t get very far. This may be expected, given the black-box and data-driven nature of these methods.<\/p>\n<p>This is another study, this time testing larger tabular datasets. By comparing 19 methods on 176 datasets, this paper shows that \ud835\uddf3\ud835\uddfc\ud835\uddff \ud835\uddee \ud835\uddf9\ud835\uddee\ud835\uddff\ud835\uddf4\ud835\uddf2 \ud835\uddfb\ud835\ude02\ud835\uddfa\ud835\uddef\ud835\uddf2\ud835\uddff \ud835\uddfc\ud835\uddf3 \ud835\uddf1\ud835\uddee\ud835\ude01\ud835\uddee\ud835\ude00\ud835\uddf2\ud835\ude01\ud835\ude00, \ud835\uddf2\ud835\uddf6\ud835\ude01\ud835\uddf5\ud835\uddf2\ud835\uddff \ud835\uddee \ud835\ude00\ud835\uddf6\ud835\uddfa\ud835\uddfd\ud835\uddf9\ud835\uddf2 \ud835\uddef\ud835\uddee\ud835\ude00\ud835\uddf2\ud835\uddf9\ud835\uddf6\ud835\uddfb\ud835\uddf2 \ud835\uddfa\ud835\uddf2\ud835\ude01\ud835\uddf5\ud835\uddfc\ud835\uddf1 \ud835\uddfd\ud835\uddf2\ud835\uddff\ud835\uddf3\ud835\uddfc\ud835\uddff\ud835\uddfa\ud835\ude00 \ud835\uddee\ud835\ude00 \ud835\ude04\ud835\uddf2\ud835\uddf9\ud835\uddf9 \ud835\uddee\ud835\ude00 \ud835\uddee\ud835\uddfb\ud835\ude06 \ud835\uddfc\ud835\ude01\ud835\uddf5\ud835\uddf2\ud835\uddff \ud835\uddfa\ud835\uddf2\ud835\ude01\ud835\uddf5\ud835\uddfc\ud835\uddf1, \ud835\uddfc\ud835\uddff \ud835\uddef\ud835\uddee\ud835\ude00\ud835\uddf6\ud835\uddf0 \ud835\uddf5\ud835\ude06\ud835\uddfd\ud835\uddf2\ud835\uddff\ud835\uddfd\ud835\uddee\ud835\uddff\ud835\uddee\ud835\uddfa\ud835\uddf2\ud835\ude01\ud835\uddf2\ud835\uddff \ud835\ude01\ud835\ude02\ud835\uddfb\ud835\uddf6\ud835\uddfb\ud835\uddf4 \ud835\uddfc\ud835\uddfb \ud835\uddee \ud835\ude01\ud835\uddff\ud835\uddf2\ud835\uddf2-\ud835\uddef\ud835\uddee\ud835\ude00\ud835\uddf2\ud835\uddf1 \ud835\uddf2\ud835\uddfb\ud835\ude00\ud835\uddf2\ud835\uddfa\ud835\uddef\ud835\uddf9\ud835\uddf2 \ud835\uddfa\ud835\uddf2\ud835\ude01\ud835\uddf5\ud835\uddfc\ud835\uddf1 \ud835\uddf6\ud835\uddfa\ud835\uddfd\ud835\uddff\ud835\uddfc\ud835\ude03\ud835\uddf2\ud835\ude00 \ud835\uddfd\ud835\uddf2\ud835\uddff\ud835\uddf3\ud835\uddfc\ud835\uddff\ud835\uddfa\ud835\uddee\ud835\uddfb\ud835\uddf0\ud835\uddf2 \ud835\uddfa\ud835\uddfc\ud835\uddff\ud835\uddf2 \ud835\ude01\ud835\uddf5\ud835\uddee\ud835\uddfb \ud835\uddf0\ud835\uddf5\ud835\uddfc\ud835\uddfc\ud835\ude00\ud835\uddf6\ud835\uddfb\ud835\uddf4 \ud835\ude01\ud835\uddf5\ud835\uddf2 \ud835\uddef\ud835\uddf2\ud835\ude00\ud835\ude01 \ud835\uddee\ud835\uddf9\ud835\uddf4\ud835\uddfc\ud835\uddff\ud835\uddf6\ud835\ude01\ud835\uddf5\ud835\uddfa.<\/p>\n<p>This project also comes with a great resource. This time it comes with a ready-to-use codebase and testbed along with the paper.<\/p>\n<p><a href=\"https:\/\/github.com\/naszilla\/tabzilla\">Source<\/a><\/p>\n","protected":false},"excerpt":{"rendered":"<p>Otherwise, tree ensembles continue to outperform neural networks. The decision tree in the figure shows the winner among the top five methods. Now, the background: I explored the why of this question before, but didn&#8217;t get very far. This may be expected, given the black-box and data-driven nature of these methods. This is another study, [&hellip;]<\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"closed","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"cybocfi_hide_featured_image":"","footnotes":""},"categories":[1],"tags":[],"class_list":["post-490","post","type-post","status-publish","format-standard","hentry","category-uncategorized"],"_links":{"self":[{"href":"https:\/\/ozer.gt\/log\/wp-json\/wp\/v2\/posts\/490","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/ozer.gt\/log\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/ozer.gt\/log\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/ozer.gt\/log\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/ozer.gt\/log\/wp-json\/wp\/v2\/comments?post=490"}],"version-history":[{"count":2,"href":"https:\/\/ozer.gt\/log\/wp-json\/wp\/v2\/posts\/490\/revisions"}],"predecessor-version":[{"id":524,"href":"https:\/\/ozer.gt\/log\/wp-json\/wp\/v2\/posts\/490\/revisions\/524"}],"wp:attachment":[{"href":"https:\/\/ozer.gt\/log\/wp-json\/wp\/v2\/media?parent=490"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/ozer.gt\/log\/wp-json\/wp\/v2\/categories?post=490"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/ozer.gt\/log\/wp-json\/wp\/v2\/tags?post=490"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}