A call-theoretic characterization of good calibration is that an agent searching for to reduce a correct loss in expectation can’t enhance their final result by post-processing a wonderfully calibrated predictor. Hu and Wu (FOCS\u201924) use this to outline an approximate calibration measure known as calibration resolution loss (CDL), which measures the maximal enchancment achievable by any post-processing over any correct loss. Sadly, CDL seems to be intractable to even weakly approximate within the offline setting, given black-box entry to the predictions and labels. We propose circumventing this by proscribing consideration to structured households of post-processing capabilities Okay<\/i>. We outline the calibration resolution loss relative to Okay<\/i>, denoted CDL_{Okay<\/i><\/sub> the place we take into account all correct losses however prohibit post-processings to a structured household Okay<\/i>. We develop a complete principle of when CDL_{Okay<\/i><\/sub> is information-theoretically and computationally tractable, and use it to show each higher and decrease bounds for pure courses Okay<\/i>. Along with introducing new definitions and algorithmic strategies to the idea of calibration for resolution making, our outcomes give rigorous ensures for some broadly used recalibration procedures in machine studying.<\/p>\n}}

\n
\u2020 College of Texas at Austin<\/li>\n
\u2021 Harvard College<\/li>\n<\/ul>\n<\/div>\n\n","protected":false},"excerpt":{"rendered":"
A call-theoretic characterization of good calibration is that an agent searching for to reduce a correct loss in expectation can’t enhance their final result by post-processing a wonderfully calibrated predictor. Hu and Wu (FOCS\u201924) use this to outline an approximate calibration measure known as calibration resolution loss (CDL), which measures the maximal enchancment achievable by […]<\/p>\n","protected":false},"author":2,"featured_media":10081,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[55],"tags":[4303,2242,3489,1625],"class_list":["post-10079","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-machine-learning","tag-calibration","tag-decision","tag-efficient","tag-making"],"_links":{"self":[{"href":"https:\/\/techtrendfeed.com\/index.php?rest_route=\/wp\/v2\/posts\/10079","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/techtrendfeed.com\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/techtrendfeed.com\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/techtrendfeed.com\/index.php?rest_route=\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/techtrendfeed.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=10079"}],"version-history":[{"count":1,"href":"https:\/\/techtrendfeed.com\/index.php?rest_route=\/wp\/v2\/posts\/10079\/revisions"}],"predecessor-version":[{"id":10080,"href":"https:\/\/techtrendfeed.com\/index.php?rest_route=\/wp\/v2\/posts\/10079\/revisions\/10080"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/techtrendfeed.com\/index.php?rest_route=\/wp\/v2\/media\/10081"}],"wp:attachment":[{"href":"https:\/\/techtrendfeed.com\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=10079"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/techtrendfeed.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=10079"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/techtrendfeed.com\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=10079"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}