[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-128042-en":3,"doc-seo-128042-105":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},128042,962084931830,"Theodore","https://ap-avatar.wpscdn.com/davatar_a8503ba1806abce46bf441b54a3ca4cd",8,"Research & Report","On Principled Modeling of Inductive Bias in Machine Learning - Dissertation","The inductive bias of a learning algorithm defines the assumptions embedded in its hypotheses and determines how effectively it generalizes to unseen data. This dissertation develops principled approaches for modeling and shaping inductive bias. It offers a unified view via decomposition of regularized empirical risk minimization, yielding value-guided modeling through regularization and data-centric modeling through training-data manipulation, with theoretical analysis, algorithmic proposals, and empirical evaluations.","On Principled Modeling of Inductive Bias in Machine Learning  \nWeiyang Liu  \nDepartment of Engineering  \nUniversity of Cambridge  \nThis dissertation is submitted for the degree of Doctor of Philosophy  \nDarwin College September 2024  \nDeclaration  \nThis thesis is the result of my own work and includes nothing which is the outcome of work done in collaboration except as declared in the preface and specified in the text. It is not substantially the same as any work that has already been submitted, or is being concurrently submitted, for any degree, diploma or other qualification at the University of Cambridge or any other University or similar institution except as declared in the preface and specified in the text. It does not exceed the prescribed word limit for the relevant Degree Committee.  \nWeiyang Liu September 2024  \nAcknowledgements  \nThe past four years of pursuing my PhD have been an incredible journey, and I am deeply grateful to the many people who have made this experience truly remarkable. First and foremost, I feel extremely fortunate to have been co-supervised by Adrian Weller and Bernhard Schölkopf, both of whom have been instrumental in my academic growth. During my first year at Cambridge, it was COVID-19 pandemic, and it is Adrian’s support and encouragement that helps me to get through. In research, his insights have always been inspiring. Moreover, he has been very supportive of my project ideas, consistently encouraging me to explore them. After moving to Tübingen, Adrian has continued to provide great support, both in my personal life and research endeavors. During my three years at Tübingen, I gained invaluable insights from Bernhard’s research vision and scientific taste. During our resarch meetings, he consistently shows a deep and insightful understanding of the field, which has significantly shaped my own perspective as a researcher. Without the guidance of both my supervisors, I could not have achieved what I have accomplished today. I am also deeply thankful for my advisor Richard Turner, who gave me a lot of useful suggestions for my research during my first year at Cambridge.  \nWhile research can often be a solitary endeavor, I have been incredibly fortunate to have many inspiring collaborators. From the Cambridge side, I want to thank Hanchen Wang, Umang Bhatt, Juyeon Heo, Katie Collins and Vihari Piratla. I have learned a lot for the collaboration with them. I still miss the good old days when I took casual walks with Hanchen and Umang. From the Tübingen side, I want to thank Yao Feng, Yandong Wen, Zhen Liu, Tim Z. Xiao, Yuliang Xiu, Zeju Qiu, Gege Gao, Haiwen Feng, Yuxuan Xue, Songyou Peng, Yucen Luo and Hongwei Yi. With all my wonderful collaborators, I have enjoyed every minute of our work together. In particular, Yao’s unwavering dedication to scientific inquiry exemplifies the true spirit of research, and she has been an constant source of inspiration. Last but not the least, I also want to thank Longhui Yu, Rongmei Lin, Shengchao Liu, Hanlin Zhang, Xiaofeng Cao and Han Shi. I cherish every  \nmoment of our collaborations. I am grateful for everyone in my lab at both Cambridge and Tübingen. My deepest gratitude extends to all my friends who have supported me throughout this long journey. Their unwavering encouragement has been invaluable, and I am forever grateful for their presence in my life.  \nMy research would not have been possible without the guidance of my other mentors, including Michael J. Black, Bhiksha Raj, and Joshua B. Tenenbaum. From them, I have gained not only knowledge but also invaluable insights into conducting world-class scientific research.  \nLastly, and most importantly of all, I am deeply grateful to my family and loved ones. My parents, Wudong Liu and Lizhen Lai, have given me unconditional support, both emotionally and financially. It is their constant encouragement that helps me discover my passion and sets me on this exciting path of inquiry.  \nAbstract  \nThe ind","cbCaimiKboBN1WkT","https://ap.wps.com/l/cbCaimiKboBN1WkT","pdf",44991134,3,1,244,"English","en",105,"# Abstract\n## Value-guided modeling through regularization\n## Data-centric modeling through training data manipulation\n## Iterative machine teaching framework\n## Open problems","[{\"question\":\"What does inductive bias mean in this thesis?\",\"answer\":\"Inductive bias is the set of assumptions used by a hypothesis to predict unseen data, which governs generalization power.\"},{\"question\":\"How does the thesis structure inductive-bias modeling?\",\"answer\":\"It proposes two perspectives: value-guided modeling via regularization and data-centric modeling via training data manipulation, motivated by decomposing regularized empirical risk minimization.\"},{\"question\":\"What is the iterative machine teaching framework?\",\"answer\":\"It studies how manipulating training data affects inductive bias, considering label synthesis teaching and data hallucination teaching, with proofs of faster convergence than a random teacher.\"}]","On Principled Modeling of Inductive Bias in Machine Learning - Dissertation | PDF",1785944363,615,{"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},"on-principled-modeling-of-inductive-bias-in-machine-learning-dissertation","",{"@graph":37,"@context":86},[38,54,69],{"@type":39,"itemListElement":40},"BreadcrumbList",[41,45,49,51],{"item":42,"name":43,"@type":44,"position":21},"https://docshare.wps.com","Home","ListItem",{"item":46,"name":47,"@type":44,"position":48},"https://docshare.wps.com/document/","Document",2,{"item":50,"name":12,"@type":44,"position":20},"https://docshare.wps.com/document/research-report/",{"item":52,"name":13,"@type":44,"position":53},"https://docshare.wps.com/document/on-principled-modeling-of-inductive-bias-in-machine-learning-dissertation/128042/",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","2026-08-05",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 does inductive bias mean in this thesis?","Question",{"text":76,"@type":77},"Inductive bias is the set of assumptions used by a hypothesis to predict unseen data, which governs generalization power.","Answer",{"name":79,"@type":74,"acceptedAnswer":80},"How does the thesis structure inductive-bias modeling?",{"text":81,"@type":77},"It proposes two perspectives: value-guided modeling via regularization and data-centric modeling via training data manipulation, motivated by decomposing regularized empirical risk minimization.",{"name":83,"@type":74,"acceptedAnswer":84},"What is the iterative machine teaching framework?",{"text":85,"@type":77},"It studies how manipulating training data affects inductive bias, considering label synthesis teaching and data hallucination teaching, with proofs of faster convergence than a random teacher.","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},[94,98,102,106,111,116,121,124,129,132,136],{"id":21,"doc_module":4,"doc_module_name":47,"category_name":95,"show_sort_weight":96,"slug":97},"Story & Novel",90,"story-novel",{"id":48,"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":112,"doc_module":4,"doc_module_name":47,"category_name":113,"show_sort_weight":114,"slug":115},6,"Technology",50,"technology",{"id":117,"doc_module":4,"doc_module_name":47,"category_name":118,"show_sort_weight":119,"slug":120},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":47,"category_name":12,"show_sort_weight":122,"slug":123},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":133,"doc_module":4,"doc_module_name":47,"category_name":134,"show_sort_weight":133,"slug":135},10,"Lifestyle","lifestyle",{"id":137,"doc_module":4,"doc_module_name":47,"category_name":138,"show_sort_weight":107,"slug":139},19,"General","general"]