[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-116847-en":3,"doc-seo-116847-105":30,"detail-sidebar-cat-0-en-105":91},{"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":4,"is_deleted":4,"is_public":20,"is_downloadable":20,"audit_status":20,"page_count":21,"language":22,"language_code":23,"site_id":24,"html_lang":23,"table_of_contents":25,"faqs":26,"seo_title":27,"seo_description":14,"update_tm":28,"read_time":29},116847,1099513958607,"Jiven","https://ap-avatar.wpscdn.com/avatar/100002390cf8733938c?x-image-process=image/resize,m_fixed,w_180,h_180&k=1778829742770036399",6,"Technology","Machine Learning-powered Course Allocation","提出一种基于机器学习的课程分配机制，将现有 Course Match 扩展为带有机器学习偏好采集模块的方案。该模块以迭代、异步方式生成针对每位学生定制的成对比较查询。激励方面，MLCM 在 Course Match 的“大规模策略抗诱导”性质下保持吸引力；福利方面，借助拟合真实数据的模拟器进行计算实验。结果显示，相比 Course Match，在仅使用十个比较查询的情况下，平均学生效用提升 4%-9%，最小学生效用提升 10%-21%，并讨论其在采用 Course Match 的高校中的落地可行性与试点便利性。","arXiv :2210 .00954v2 [ cs .GT] 10 Mar 2023  \nMachine Learning-powered Course Allocation  \nERMIS SOUMALIAS∗, University of Zurich & ETH AI Center, [ermis@ifi.uzh.ch](ermis@ifi.uzh.ch)[ ](ermis@ifi.uzh.ch)BEHNOOSH ZAMANLOOY∗, McMaster University, [zamanlob@mcmaster.ca](zamanlob@mcmaster.ca)[ ](zamanlob@mcmaster.ca)JAKOB WEISSTEINER, University of Zurich & ETH AI Center, [weissteiner@ifi.uzh.ch](weissteiner@ifi.uzh.ch)[ ](weissteiner@ifi.uzh.ch)SVEN SEUKEN, University of Zurich & ETH AI Center, [seuken@ifi.uzh.ch](seuken@ifi.uzh.ch)  \nWe introduce a machine learning-powered course allocation mechanism. Concretely, we extend the stateof-the-art Course Match mechanism with a machine learning-based preference elicitation module. In an iterative, asynchronous manner, this module generates pairwise comparison queries that are tailored to each individual student. Regarding incentives, our machine learning-powered course match (MLCM) mechanism retains the attractive strategyproofness in the large property of Course Match. Regarding welfare, we perform computational experiments using a simulator that was fitted to real-world data. Our results show that, compared to Course Match, MLCM increases average student utility by 4%-9% and minimum student utility by 10%-21%, even with only ten comparison queries. Finally, we highlight the practicability of MLCM and the ease of piloting it for universities currently using Course Match.  \n∗ These authors contributed equally.  \nMachine Learning-powered Course Allocation 1  \n1 INTRODUCTION  \nThe course allocation problem arises when educational institutions assign bundles of courses to students [Budish and Cantillon, 2012] . Each course has a limited number of indivisible seats and monetary transfers are prohibited for fairness reasons. What makes this problem particularly challenging is the combinatorial structure of the students’ preferences, as students may view certain courses as complements or substitutes [Budish and Kessler, 2022] .  \n1.1 Course Match  \nThe state-of-the-art practical solution to the course allocation problem is the Course Match (CM) mechanism by Budish et al. [2017], which provides a good trade-off between efficiency, fairness, and incentives. CM has now been adopted in many universities such as the Wharton School at the University of Pennsylvania and Columbia Business School.  \nCM uses a simple reporting language to elicit students’ preferences over schedules (i.e., course bundles) . Concretely, CM offers students a graphical user interface (GUI) to enter a base value between 0 and 100 for each course, and an adjustment value between −200 and 200 for each pair of courses. Adjustments allow students to report complementarities and substitutabilities between courses, up to pairwise interactions. The total value of a schedule is then the sum of the base values reported for each course in that schedule plus any adjustments (if both courses are in the schedule) . Prior to the adoption of CM in practice, Budish and Kessler [2022] performed a lab experiment to evaluate CM. Regarding efficiency, they found that, on average, students were happier with CM compared to the Bidding Points Auction [Sönmez and Ünver, 2010], the previously used mechanism.  \nRegarding fairness, students also found CM fairer. Regarding the reporting language, they found that students were able to report their preferences “accurately enough to realize CM’s theoretical benefits.” Given these positive findings, Wharton was then the first school to switch to CM.  \n1.2 Preference Elicitation Shortcomings of Course Match  \nHowever, Budish et al. [2017] were already concerned that the CM language may not be able to fully capture all students’ preferences. Furthermore, they mentioned that some students might find it non-trivial to use the CM language and might therefore make mistakes when reporting their preferences. Indeed, the lab experiment by Budish and Kessler [2022] revealed several shortcomings of CM in this rega","cbCaimBcHMX3H4gA","https://ap.wps.com/l/cbCaimBcHMX3H4gA","pdf",1134209,1,48,"English","en",105,"# Introduction\n## Course Match\n## Preference Elicitation Shortcomings of Course Match\n## Machine Learning-powered Preference Elicitation","[{\"question\":\"什么是本文件提出的机器学习驱动课程分配机制？\",\"answer\":\"文中提出一种将 Course Match 扩展的 MLCM 机制，并加入机器学习偏好采集模块，通过成对比较查询更好地捕捉学生偏好，从而分配课程。\"},{\"question\":\"MLCM 在激励与公平性方面与 Course Match 的关系是什么？\",\"answer\":\"文中指出，MLCM 在 Course Match 的“大规模策略抗诱导”性质上保留了其吸引人的激励特征。\"},{\"question\":\"实验结果表明 MLCM 相比 Course Match 带来了哪些福利改进？\",\"answer\":\"在仅使用十个比较查询的条件下，MLCM 将平均学生效用提升 4%-9%，最小学生效用提升 10%-21%，并展示了该方法的可操作性与试点难度较低。\"}]","Machine Learning-powered Course Allocation | PDF",1785672057,121,{"code":4,"msg":31,"data":32},"ok",{"site_id":24,"language":23,"slug":33,"title":13,"keywords":34,"description":14,"schema_data":35,"social_meta":86,"head_meta":88,"extra_data":90,"updated_unix":28},"machine-learning-powered-course-allocation","",{"@graph":36,"@context":85},[37,54,68],{"@type":38,"itemListElement":39},"BreadcrumbList",[40,44,48,51],{"item":41,"name":42,"@type":43,"position":20},"https://docshare.wps.com","Home","ListItem",{"item":45,"name":46,"@type":43,"position":47},"https://docshare.wps.com/document/","Document",2,{"item":49,"name":12,"@type":43,"position":50},"https://docshare.wps.com/document/technology/",3,{"item":52,"name":13,"@type":43,"position":53},"https://docshare.wps.com/document/machine-learning-powered-course-allocation/116847/",4,{"url":52,"name":13,"@type":55,"author":56,"headline":13,"publisher":58,"fileFormat":61,"inLanguage":23,"description":14,"dateModified":62,"datePublished":62,"encodingFormat":61,"isAccessibleForFree":63,"interactionStatistic":64},"DigitalDocument",{"name":9,"@type":57},"Person",{"url":41,"name":59,"@type":60},"DocShare","Organization","application/pdf","2026-08-02",true,{"@type":65,"interactionType":66,"userInteractionCount":4},"InteractionCounter",{"@type":67},"ViewAction",{"@type":69,"mainEntity":70},"FAQPage",[71,77,81],{"name":72,"@type":73,"acceptedAnswer":74},"什么是本文件提出的机器学习驱动课程分配机制？","Question",{"text":75,"@type":76},"文中提出一种将 Course Match 扩展的 MLCM 机制，并加入机器学习偏好采集模块，通过成对比较查询更好地捕捉学生偏好，从而分配课程。","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"MLCM 在激励与公平性方面与 Course Match 的关系是什么？",{"text":80,"@type":76},"文中指出，MLCM 在 Course Match 的“大规模策略抗诱导”性质上保留了其吸引人的激励特征。",{"name":82,"@type":73,"acceptedAnswer":83},"实验结果表明 MLCM 相比 Course Match 带来了哪些福利改进？",{"text":84,"@type":76},"在仅使用十个比较查询的条件下，MLCM 将平均学生效用提升 4%-9%，最小学生效用提升 10%-21%，并展示了该方法的可操作性与试点难度较低。","https://schema.org",{"og:url":52,"og:type":87,"og:title":13,"og:site_name":59,"og:description":14},"article",{"robots":89,"canonical":52},"index,follow",{"doc_id":7,"site_id":24},{"code":4,"msg":5,"data":92},[93,97,101,105,110,113,118,123,128,131,135],{"id":20,"doc_module":4,"doc_module_name":46,"category_name":94,"show_sort_weight":95,"slug":96},"Story & Novel",90,"story-novel",{"id":47,"doc_module":4,"doc_module_name":46,"category_name":98,"show_sort_weight":99,"slug":100},"Literature",80,"literature",{"id":53,"doc_module":4,"doc_module_name":46,"category_name":102,"show_sort_weight":103,"slug":104},"Exam",70,"exam",{"id":106,"doc_module":4,"doc_module_name":46,"category_name":107,"show_sort_weight":108,"slug":109},5,"Comic",60,"comic",{"id":11,"doc_module":4,"doc_module_name":46,"category_name":12,"show_sort_weight":111,"slug":112},50,"technology",{"id":114,"doc_module":4,"doc_module_name":46,"category_name":115,"show_sort_weight":116,"slug":117},7,"Healthcare",40,"healthcare",{"id":119,"doc_module":4,"doc_module_name":46,"category_name":120,"show_sort_weight":121,"slug":122},8,"Research & Report",30,"research-report",{"id":124,"doc_module":4,"doc_module_name":46,"category_name":125,"show_sort_weight":126,"slug":127},9,"Religion & Spirituality",20,"religion-spirituality",{"id":126,"doc_module":4,"doc_module_name":46,"category_name":129,"show_sort_weight":126,"slug":130},"World Cup","world-cup",{"id":132,"doc_module":4,"doc_module_name":46,"category_name":133,"show_sort_weight":132,"slug":134},10,"Lifestyle","lifestyle",{"id":136,"doc_module":4,"doc_module_name":46,"category_name":137,"show_sort_weight":106,"slug":138},19,"General","general"]