[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-120564-en":3,"doc-seo-120564-105":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},120564,687197207919,"Theodora","https://ap-avatar.wpscdn.com/avatar/a000253d6f5f7c60be?x-image-process=image/resize,m_fixed,w_180,h_180&k=1779446848396160552",8,"Research & Report","STATISTICAL MACHINE LEARNING BEYOND STANDARD SUPERVISED LEARNING - Dissertation","A dissertation presenting statistical learning methods for challenges that standard supervised learning does not address, including distribution shifts, incomplete supervision, and corrupted or missing labels. The work studies four learning problems outside the conventional setting and extends foundational statistical learning principles to obtain generalizable, robust algorithms with theoretical guarantees. It develops a class probability matching framework for label shift in transfer learning and proposes methods for partial label learning using leveraged weighted loss, supported by risk consistency results and empirical validation.","STATISTICAL MACHINE LEARNING BEYOND STANDARD SUPERVISED LEARNING  \nHongwei Wen  \nSTATISTICAL MACHINE LEARNING BEYOND STANDARD SUPERVISED LEARNING  \nDISSERTATION  \nto obtain  \nthe degree of doctor at the University of Twente, on the authority of the rector magnificus, [prof. dr. ir. A. Veldkamp](prof. dr. ir. A. Veldkamp), on account of the decision of the Doctorate Board, to be publicly defended on Tuesday 9 September 2025 at 10.45 hours  \nby  \nHongwei Wen  \nborn on the 2nd of April, 1996  \nin Tianjin, China  \nThis dissertation has been approved by:  \nPromotors:  \nprof. dr. A.J. Schmidt-Hieber  \nprof. dr. W.M. Koolen  \nCo-promotor: dr. A. Betken  \nCover design: Hongwei Wen  \nPrinted by: Ipskamp Printing  \nLay-out: Hongwei Wen  \nISBN (print): 978-90-365-6819-7  \nISBN (digital): 978-90-365-6820-3  \nURL: [https://doi.org/10.3990/1.9789036568203](https://doi.org/10.3990/1.9789036568203)  \n© 2025 Hongwei Wen, The Netherlands. All rights reserved. No parts of this thesis may be reproduced, stored in a retrieval system or transmitted in any form or by any means without permission of the author. Alle rechten voorbehouden. Niets uit deze uitgave mag worden vermenigvuldigd, in enige vorm of op enige wijze, zonder voorafgaande schriftelijke toestemming van de auteur.  \nGRADUATION COMMITTEE:  \nChairman/secretary prof.dr.ir. B.R.H.M. Haverkort  \nPromotor prof. dr. A.J. Schmidt-Hieber  \nUniversiteit Twente, EEMCS, Mathematics of Operations Research prof. dr. W.M. Koolen  \nUniversiteit Twente, EEMCS, Mathematics of Operations Research  \nCo-promotor dr. A. Betken  \nUniversiteit Twente, EEMCS, Mathematics of Operations Research  \nMembers prof. dr. C. Brune  \nUniversiteit Twente, EEMCS, Mathematics of Imaging & AIdr. J.M. Wolterink  \nUniversiteit Twente, EEMCS, Mathematics of Imaging & AI prof. dr. N. M¨ucke  \nTechnische Universit¨at Carolo-Wilhelmina Braunschweig, Carl-Friedrich-Gau¨b-Fakult¨at  \nprof. dr. W. Wang  \nUniversity of Groningen (will soon move to University of Bristol), Faculty of Economics and Business  \nAcknowledgements  \nThe completion of my PhD has been an incredible journey, made possible by the guidance, encouragement, and support I received throughout my time with the Statistics research group. I would like to take this opportunity to sincerely thank everyone who has supported and contributed to this journey in various ways.  \nFirst, I am deeply grateful to my daily supervisor, Annika Betken. Thankyou for the time, energy, and care you dedicated to our weekly discussions and research projects. Your warmth and kindness—not only in academic matters but also in everyday life—have accompanied and supported me throughout this journey. I would also like to extend my sincere thanks to my promotor, Johannes Schmidt-Hieber. Your thoughtful feedback and insightful discussions have been invaluable in shaping this thesis. Your high standards and rich research experience have inspired me and motivated me to become a more rigorous and wellrounded researcher. My sincere appreciation also goes to Wouter Koolen, who introduced me to the world of multi-armed bandits. Our discussions have always been intellectually stimulating and full of fresh ideas. I have learned tremendously from you. In addition, I would like to thank Hans Hang for his careful guidance and innovative perspectives in our collaborative work.  \nI am also grateful to my fellow PhD students and postdocs. I have thoroughly enjoyed our conversations and collaborations. Beyond the academic environment, I deeply cherish the friendships I have formed in Enschede—thank you for making this time memorable. Finally, I would like to thank my family for their unwavering support and understanding throughout this journey. Your belief in me has been my greatest strength.  \nSummary  \nMachine learning has become a pivotal technology, driving advancements across scientific research, industry, and society. Traditional supervised learning delivered substantial progress in solving problems within","cbCainHqFF22pGnJ","https://ap.wps.com/l/cbCainHqFF22pGnJ","pdf",4405418,1,257,"English","en",105,"# Acknowledgements\n# Summary\n## Label shift in transfer learning\n## Partial label learning","[{\"question\":\"What limitations of standard supervised learning motivate this dissertation?\",\"answer\":\"Real-world tasks often face distribution shifts, incomplete supervision, and corrupted or missing labels, which fall outside the standard supervised learning assumptions.\"},{\"question\":\"How does the thesis address label shift in transfer learning?\",\"answer\":\"It introduces class probability matching (CPM), which aligns class probabilities from the source domain with weighted class probabilities of target samples, and it proposes calibrated neural network and kernel logistic regression variants with theoretical guarantees.\"},{\"question\":\"What is leveraged weighted (LW) loss and what does it achieve?\",\"answer\":\"LW loss is designed for partial label learning by introducing a leverage parameter β to balance losses from partially and fully observed labels. The dissertation establishes risk consistency properties and confirms strong empirical performance on real-world data.\"}]","STATISTICAL MACHINE LEARNING BEYOND STANDARD SUPERVISED LEARNING - Dissertation | PDF",1785730665,648,{"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},"statistical-machine-learning-beyond-standard-supervised-learning-dissertation","",{"@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/research-report/",3,{"item":52,"name":13,"@type":43,"position":53},"https://docshare.wps.com/document/statistical-machine-learning-beyond-standard-supervised-learning-dissertation/120564/",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","2026-08-03",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},"What limitations of standard supervised learning motivate this dissertation?","Question",{"text":75,"@type":76},"Real-world tasks often face distribution shifts, incomplete supervision, and corrupted or missing labels, which fall outside the standard supervised learning assumptions.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"How does the thesis address label shift in transfer learning?",{"text":80,"@type":76},"It introduces class probability matching (CPM), which aligns class probabilities from the source domain with weighted class probabilities of target samples, and it proposes calibrated neural network and kernel logistic regression variants with theoretical guarantees.",{"name":82,"@type":73,"acceptedAnswer":83},"What is leveraged weighted (LW) loss and what does it achieve?",{"text":84,"@type":76},"LW loss is designed for partial label learning by introducing a leverage parameter β to balance losses from partially and fully observed labels. The dissertation establishes risk consistency properties and confirms strong empirical performance on real-world data.","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},[93,97,101,105,110,115,120,123,128,131,135],{"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":111,"doc_module":4,"doc_module_name":46,"category_name":112,"show_sort_weight":113,"slug":114},6,"Technology",50,"technology",{"id":116,"doc_module":4,"doc_module_name":46,"category_name":117,"show_sort_weight":118,"slug":119},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":46,"category_name":12,"show_sort_weight":121,"slug":122},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"]