[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-128703-en":3,"doc-seo-128703-105":31,"detail-sidebar-cat-0-en-105":92},{"code":4,"msg":5,"data":6},0,"success",{"doc_id":7,"user_id":8,"nickname":9,"user_avatar":10,"doc_module":4,"category_id":11,"category_name":12,"doc_title":13,"doc_description":14,"doc_content":15,"file_id":16,"file_url":17,"file_type":18,"file_size":19,"view_count":20,"is_deleted":4,"is_public":21,"is_downloadable":21,"audit_status":21,"page_count":22,"language":23,"language_code":24,"site_id":25,"html_lang":24,"table_of_contents":26,"faqs":27,"seo_title":28,"seo_description":14,"update_tm":29,"read_time":30},128703,962084926284,"Aurora","https://ap-avatar.wpscdn.com/davatar_29158cc5080c5b710cf443261637dec0",6,"Technology","Just Sort It! - A Simple and Effective Approach to Active Preference Learning","Active preference learning is studied through adaptively selected pairwise comparisons, aiming to recover an accurate ranking while using comparisons sparingly. When outcomes are consistent with the target order, efficient sorting methods like Quicksort are optimal, but the behavior under inconsistencies is harder. The work provides favorable guarantees for Quicksort under the Bradley–Terry model, then shows experimentally that repeatedly sorting items yields an active learning strategy matching state-of-the-art accuracy with far lower computational cost than random sampling.","View metadata, citation and similar [papers at ](papers at core.ac.uk)[core.ac.uk](papers at core.ac.uk) brought to you by CORE  \nprovided by Infoscience- École polytechnique fédérale de Lausanne  \nJust Sort It! A Simple and Effective Approach to Active Preference Learning  \nLucas Maystre 1 Matthias Grossglauser 1  \nAbstract  \nWe address the problem of learning a ranking by using adaptively chosen pairwise comparisons.  \nOur goal is to recover the ranking accurately but to sample the comparisons sparingly. If all comparison outcomes are consistent with the ranking, the optimal solution is to use an ef􀀂cient sorting algorithm, such as Quicksort. But how do sorting algorithms behave if some comparison outcomes are inconsistent with the ranking? We give favorable guarantees for Quicksort for the popular Bradley–Terry model, under natural assumptionson the parameters. Furthermore, we empirically demonstrate that sorting algorithms lead to a very simple and effective active learning strategy: repeatedly sort the items. This strategy performs as well as state-of-the-art methods (and much better than random sampling) at a minuscule fraction of the computational cost.  \n1 Introduction  \nThe problem of recovering a ranking over n items from noisy outcomes of pairwise comparisons has attracted, in the last century, much research interest, driven by applications in sports (Elo, 1978), social sciences (Thurstone, 1927; Salganik & Levy, 2015) and—more recently—recommender systems (Houlsby et al., 2012) . Whereas pairwise comparison models and related inference algorithms have been extensively studied, the issue of which pairwise comparisons to sample, also known as active learning, has received signi􀀂cantly less attention. To understand the potential bene-􀀂ts of adaptively selecting samples, consider the case where comparison outcomes are noiseless, i.e., consistent with alinear order on a set of n items. If pairs of items are selected at random, it is necessary to collect 􀀊(n2 ) comparisons to recover the ranking (Alon et al., 1994) . In contrast, by using an ef􀀂cient sorting algorithm, O(nlog n) adaptively  \n1 School of Computer and Communication Sciences, EPFL, Lausanne, Switzerland. Correspondence to: Lucas Maystre \u003Clucas.maystre@ep􀀃.ch> .  \nProceedings of the 34 th International Conference on Machine Learning, Sydney, Australia, PMLR 70, 2017. Copyright 2017 by the author(s) .  \nchosen comparisons are suf􀀂cient. In this work, we demonstrate that sorting algorithms can also be helpful in the noisy setting, where some comparison outcomes are inconsistent with the ranking: despite errors, sorting algorithms tend to select informative samples. We focus on the Bradley–Terry (BT) model, a widely-used probabilistic model of comparison outcomes. In this model, each item is associated with a parameter on the real line, and the probability of observing an incorrect outcome decreases as the distance between the items’ parameters increases.  \nFirst, we study the output of a single execution of Quicksort when comparison outcomes are generated from a BT model, under the assumption that the distance between adjacent parameters is (stochastically) uniform across the ranking. We measure the quality of a ranking estimate by its displacement with respect to the ground truth, i.e., the sum of rank differences. We show that Quicksort’s output is a good approximation to the ground-truth ranking: no method comparing every pair of items at most once can do better (up to constant factors). Furthermore, we show that by aggregating O(log5 n) independent runs of Quicksort, it is possible to recover the exact rank for all but a vanishing fraction of the items. These theoretical results suggest that adaptive sampling is able to bring a substantial acceleration to the learning process.  \nSecond, we propose a practical active-learning (AL) strategy that consists of repeatedly sorting the items. We evaluate our sorting-based method on three datasets and compare it to","cbCaiu47aRWSgk1a","https://ap.wps.com/l/cbCaiu47aRWSgk1a","pdf",254353,2,1,10,"English","en",105,"# Introduction\n## Problem setting and motivation\n## Noiseless versus noisy comparisons\n## Contributions and overview\n# Preliminaries and Notation\n## Items, comparisons, and consistency\n## Bradley–Terry model\n## Ranking metrics and displacement","[{\"question\":\"What problem does the paper address in active preference learning?\",\"answer\":\"It targets learning an unknown ranking by adaptively selecting pairwise comparisons while minimizing how many comparisons are needed.\"},{\"question\":\"How does the paper analyze Quicksort under noisy comparisons?\",\"answer\":\"It studies Quicksort’s output when comparison outcomes follow the Bradley–Terry model and derives guarantees about ranking recovery quality despite inconsistencies.\"},{\"question\":\"What is the proposed practical active-learning strategy?\",\"answer\":\"The strategy repeatedly sorts the items and aggregates the results, achieving ranking estimates that perform like state-of-the-art methods at a much lower computational cost than random sampling.\"}]","Just Sort It! - A Simple and Effective Approach to Active Preference Learning | PDF",1786002761,25,{"code":4,"msg":32,"data":33},"ok",{"site_id":25,"language":24,"slug":34,"title":13,"keywords":35,"description":14,"schema_data":36,"social_meta":87,"head_meta":89,"extra_data":91,"updated_unix":29},"just-sort-it-a-simple-and-effective-approach-to-active-preference-learning","",{"@graph":37,"@context":86},[38,54,69],{"@type":39,"itemListElement":40},"BreadcrumbList",[41,45,48,51],{"item":42,"name":43,"@type":44,"position":21},"https://docshare.wps.com","Home","ListItem",{"item":46,"name":47,"@type":44,"position":20},"https://docshare.wps.com/document/","Document",{"item":49,"name":12,"@type":44,"position":50},"https://docshare.wps.com/document/technology/",3,{"item":52,"name":13,"@type":44,"position":53},"https://docshare.wps.com/document/just-sort-it-a-simple-and-effective-approach-to-active-preference-learning/128703/",4,{"url":52,"name":13,"@type":55,"author":56,"headline":13,"publisher":58,"fileFormat":61,"inLanguage":24,"description":14,"dateModified":62,"datePublished":63,"encodingFormat":61,"isAccessibleForFree":64,"interactionStatistic":65},"DigitalDocument",{"name":9,"@type":57},"Person",{"url":42,"name":59,"@type":60},"DocShare","Organization","application/pdf","2026-08-25","2026-08-06",true,{"@type":66,"interactionType":67,"userInteractionCount":20},"InteractionCounter",{"@type":68},"ViewAction",{"@type":70,"mainEntity":71},"FAQPage",[72,78,82],{"name":73,"@type":74,"acceptedAnswer":75},"What problem does the paper address in active preference learning?","Question",{"text":76,"@type":77},"It targets learning an unknown ranking by adaptively selecting pairwise comparisons while minimizing how many comparisons are needed.","Answer",{"name":79,"@type":74,"acceptedAnswer":80},"How does the paper analyze Quicksort under noisy comparisons?",{"text":81,"@type":77},"It studies Quicksort’s output when comparison outcomes follow the Bradley–Terry model and derives guarantees about ranking recovery quality despite inconsistencies.",{"name":83,"@type":74,"acceptedAnswer":84},"What is the proposed practical active-learning strategy?",{"text":85,"@type":77},"The strategy repeatedly sorts the items and aggregates the results, achieving ranking estimates that perform like state-of-the-art methods at a much lower computational cost than random sampling.","https://schema.org",{"og:url":52,"og:type":88,"og:title":13,"og:site_name":59,"og:description":14},"article",{"robots":90,"canonical":52},"index,follow",{"doc_id":7,"site_id":25},{"code":4,"msg":5,"data":93},[94,98,102,106,111,114,119,124,129,132,135],{"id":21,"doc_module":4,"doc_module_name":47,"category_name":95,"show_sort_weight":96,"slug":97},"Story & Novel",90,"story-novel",{"id":20,"doc_module":4,"doc_module_name":47,"category_name":99,"show_sort_weight":100,"slug":101},"Literature",80,"literature",{"id":53,"doc_module":4,"doc_module_name":47,"category_name":103,"show_sort_weight":104,"slug":105},"Exam",70,"exam",{"id":107,"doc_module":4,"doc_module_name":47,"category_name":108,"show_sort_weight":109,"slug":110},5,"Comic",60,"comic",{"id":11,"doc_module":4,"doc_module_name":47,"category_name":12,"show_sort_weight":112,"slug":113},50,"technology",{"id":115,"doc_module":4,"doc_module_name":47,"category_name":116,"show_sort_weight":117,"slug":118},7,"Healthcare",40,"healthcare",{"id":120,"doc_module":4,"doc_module_name":47,"category_name":121,"show_sort_weight":122,"slug":123},8,"Research & Report",30,"research-report",{"id":125,"doc_module":4,"doc_module_name":47,"category_name":126,"show_sort_weight":127,"slug":128},9,"Religion & Spirituality",20,"religion-spirituality",{"id":127,"doc_module":4,"doc_module_name":47,"category_name":130,"show_sort_weight":127,"slug":131},"World Cup","world-cup",{"id":22,"doc_module":4,"doc_module_name":47,"category_name":133,"show_sort_weight":22,"slug":134},"Lifestyle","lifestyle",{"id":136,"doc_module":4,"doc_module_name":47,"category_name":137,"show_sort_weight":107,"slug":138},19,"General","general"]