[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-seo-290726-105":3,"detail-sidebar-cat-0-en-105":81,"doc-detail-290726-en":130},{"code":4,"msg":5,"data":6},0,"ok",{"site_id":7,"language":8,"slug":9,"title":10,"keywords":11,"description":12,"schema_data":13,"social_meta":74,"head_meta":76,"extra_data":78,"updated_unix":80},105,"en","algorithmic-fairness-in-machine-learning-metrics-and-algorithms-dissertation","ALGORITHMIC FAIRNESS IN MACHINE LEARNING: METRICS AND ALGORITHMS - Dissertation","","The widespread use of machine learning in real applications has intensified concerns about fairness, especially in sensitive settings involving life-changing decisions. This dissertation targets discriminatory behavior toward populations and demographic groups, showing that fairness requires appropriate fairness metrics and debiasing of data and models. It expands fairness beyond metric, data, and model design: proposing GDP for tractable continuous/categorical group fairness, analyzing bias from graph aggregation with FairGR, developing fair message passing with unified optimization, and introducing RFR for worst-case robustness under distribution shifts.",{"@graph":14,"@context":73},[15,34,56],{"@type":16,"itemListElement":17},"BreadcrumbList",[18,23,27,31],{"item":19,"name":20,"@type":21,"position":22},"https://docshare.wps.com","Home","ListItem",1,{"item":24,"name":25,"@type":21,"position":26},"https://docshare.wps.com/document/","Document",2,{"item":28,"name":29,"@type":21,"position":30},"https://docshare.wps.com/document/research-report/","Research & Report",3,{"item":32,"name":10,"@type":21,"position":33},"https://docshare.wps.com/document/algorithmic-fairness-in-machine-learning-metrics-and-algorithms-dissertation/290726/",4,{"url":32,"name":10,"@type":35,"image":36,"author":41,"headline":10,"publisher":44,"fileFormat":47,"inLanguage":8,"description":12,"dateModified":48,"datePublished":49,"encodingFormat":47,"isAccessibleForFree":50,"interactionStatistic":51},"DigitalDocument",{"url":37,"@type":38,"width":39,"height":40},"https://docshare.wps.com/thumbnails/algorithmic-fairness-in-machine-learning-metrics-and-algorithms-dissertation/290726.png","ImageObject",300,407,{"name":42,"@type":43},"Miles","Person",{"url":19,"name":45,"@type":46},"DocShare","Organization","application/pdf","2026-10-29","2026-09-17",true,{"@type":52,"interactionType":53,"userInteractionCount":55},"InteractionCounter",{"@type":54},"ViewAction",8,{"@type":57,"mainEntity":58},"FAQPage",[59,65,69],{"name":60,"@type":61,"acceptedAnswer":62},"Why is algorithmic fairness important in machine learning?","Question",{"text":63,"@type":64},"Because machine learning used in sensitive, high-stakes decisions can produce discriminatory behavior toward specific populations or demographic groups.","Answer",{"name":66,"@type":61,"acceptedAnswer":67},"What is GDP in this dissertation?",{"text":68,"@type":64},"Generalized Demographic Parity (GDP) is a newly designed group fairness metric for continuous and categorical sensitive attributes with tractable computations and reliability.",{"name":70,"@type":61,"acceptedAnswer":71},"How does the dissertation address fairness for graph neural networks?",{"text":72,"@type":64},"It studies why aggregation can increase bias in graph data, motivates the Fair Graph Rewiring (FairGR) algorithm, and develops a Fair Message Passing (FMP) approach that achieves fairness as aggregation followed by debiasing.","https://schema.org",{"og:url":32,"og:type":75,"og:title":10,"og:site_name":45,"og:description":12},"article",{"robots":77,"canonical":32},"index,follow",{"doc_id":79,"site_id":7},290726,1789642449,{"code":4,"msg":82,"data":83},"success",[84,88,92,96,101,106,111,114,119,122,126],{"id":22,"doc_module":4,"doc_module_name":25,"category_name":85,"show_sort_weight":86,"slug":87},"Story & Novel",90,"story-novel",{"id":26,"doc_module":4,"doc_module_name":25,"category_name":89,"show_sort_weight":90,"slug":91},"Literature",80,"literature",{"id":33,"doc_module":4,"doc_module_name":25,"category_name":93,"show_sort_weight":94,"slug":95},"Exam",70,"exam",{"id":97,"doc_module":4,"doc_module_name":25,"category_name":98,"show_sort_weight":99,"slug":100},5,"Comic",60,"comic",{"id":102,"doc_module":4,"doc_module_name":25,"category_name":103,"show_sort_weight":104,"slug":105},6,"Technology",50,"technology",{"id":107,"doc_module":4,"doc_module_name":25,"category_name":108,"show_sort_weight":109,"slug":110},7,"Healthcare",40,"healthcare",{"id":55,"doc_module":4,"doc_module_name":25,"category_name":29,"show_sort_weight":112,"slug":113},30,"research-report",{"id":115,"doc_module":4,"doc_module_name":25,"category_name":116,"show_sort_weight":117,"slug":118},9,"Religion & Spirituality",20,"religion-spirituality",{"id":117,"doc_module":4,"doc_module_name":25,"category_name":120,"show_sort_weight":117,"slug":121},"World Cup","world-cup",{"id":123,"doc_module":4,"doc_module_name":25,"category_name":124,"show_sort_weight":123,"slug":125},10,"Lifestyle","lifestyle",{"id":127,"doc_module":4,"doc_module_name":25,"category_name":128,"show_sort_weight":97,"slug":129},19,"General","general",{"code":4,"msg":82,"data":131},{"doc_id":79,"user_id":132,"nickname":42,"user_avatar":133,"doc_module":4,"category_id":55,"category_name":29,"doc_title":10,"doc_description":12,"doc_content":134,"file_id":135,"file_url":136,"file_type":137,"file_size":138,"view_count":55,"is_deleted":4,"is_public":22,"is_downloadable":22,"audit_status":22,"page_count":139,"language":140,"language_code":8,"site_id":7,"html_lang":8,"table_of_contents":141,"faqs":142,"seo_title":143,"seo_description":12,"update_tm":80,"read_time":39,"preview_page_count":4},13056703019404,"https://ap-avatar.wpscdn.com/davatar_29158cc5080c5b710cf443261637dec0","ALGORITHMIC FAIRNESS IN MACHINE LEARNING: METRICS AND ALGORITHMS  \nA Dissertation  \nby  \nZHIMENG JIANG  \nSubmitted to the Graduate and Professional School of  \nTexas A&M University  \nin partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY  \nChair of Committee, Xia (Ben) Hu Committee Members, Shuiwang Ji  \nRuihong Huang  \nAli Mostafavi  \nHead of Department, Scott Schaefer  \nDecember 2023  \nMajor Subject: Computer Science  \nCopyright 2023 Zhimeng Jiang  \nABSTRACT  \nThe widespread use of machine learning in various applications has raised concerns about ensuring fairness, particularly in sensitive scenarios where life-changing decisions are made. It is crucial to prevent machine learning algorithms from exhibiting discriminatory behaviors towards certain populations or demographic groups and achieving fairness involves designing appropriate fairness metrics and debiasing data and models. However, existing fairness work has limited application scope toward achieving fairness.  \nTo this end, we aim to extend the application scope from fairness metric, data, and model perspectives. Firstly, existing group fairness metrics for continuous sensitive attributes are intractable due to involved functional optimization. We design a new group fairness metric, named Generalized Demographic Parity (GDP), for continuous and categorical sensitive attributes with tractable computations and reliability. Secondly, in graph learning, graph neural networks (GNNs) exhibit higher prediction bias than multi-layer perceptron (MLP) due to the aggregation operation without theoretical understanding. We provide a theoretical study to interpret when and why aggregation enhances bias from a graph data perspective, motivating a Fair Graph Rewiring (FairGR) algorithm. Thirdly, we aim to develop a Fair Message Passing (FMP) approach to achieve fairness in graph from a model perspective. We consider fairness objectives in the unified optimization framework for many GNNs and develop a solver to obtain fair message passing, which is interpreted as aggregation followed by debiasing. Finally, to enhance fairness under distribution shifts problem, we propose fair robust fairness regularization (RFR) by considering the worst case within the weight perturbation ball for each sensitive attribute group. In this dissertation, we made a series of contributions to enhance fairness in machine learning from various (e.g., fairness metrics, data, and model) perspectives.  \nDEDICATION  \nTo my mother, father, grandfather, and grandmother.  \nACKNOWLEDGMENTS  \nThis dissertation would have never been completed without a few great people who offered unwavering support, inspiration, and guidance.  \nI am deeply grateful to my advisor, Dr. Xia (Ben) Hu, whose mentorship, wisdom, and encouragement were invaluable. His dedication to research, his students, and his community has profoundly influenced me, shaping me into not just a better researcher but also a better person. I am equally thankful to my dissertation committee members -Dr. Shuiwang Ji, Dr. Ruihong Huang, and Dr. Ali Mostafavi-for their insightful feedback and consistent support throughout my Ph.D. journey.  \nIn addition, I’d like to express my gratitude to Dr. Na Zou and Dr. Chao Fan for their unwavering support and guidance. Collaborating with them has been insightful and enlightening. I’m also thankful to Dr. Rui Chen and Dr. Li Li for offering me my first industry (remote) internship at Samsung Research America and for their help. My time at Visa Research, enhanced by the guidance of Dr. Huiyuan Chen, Dr. Menghai Pan, Dr. Xiaoting Li, and Dr. Hao Yang, provided me with unparalleled freedom and enriched my research journey with stimulating discussions. I’m grateful to Dr. Xianfeng Tang, Dr. Haoming Jiang, and Mr. Jingfeng Yang for their assistance during my internship at Amazon. My sincere thanks to my colleagues at DATA Lab for their unwavering support and invaluable discussions. Our time spent s","cbCaihP9Ybjn9ZPz","https://ap.wps.com/l/cbCaihP9Ybjn9ZPz","pdf",4155348,119,"English","# ABSTRACT\n# DEDICATION\n# ACKNOWLEDGMENTS\n# CONTRIBUTORS AND FUNDING SOURCES\n# TABLE OF CONTENTS","[{\"question\":\"Why is algorithmic fairness important in machine learning?\",\"answer\":\"Because machine learning used in sensitive, high-stakes decisions can produce discriminatory behavior toward specific populations or demographic groups.\"},{\"question\":\"What is GDP in this dissertation?\",\"answer\":\"Generalized Demographic Parity (GDP) is a newly designed group fairness metric for continuous and categorical sensitive attributes with tractable computations and reliability.\"},{\"question\":\"How does the dissertation address fairness for graph neural networks?\",\"answer\":\"It studies why aggregation can increase bias in graph data, motivates the Fair Graph Rewiring (FairGR) algorithm, and develops a Fair Message Passing (FMP) approach that achieves fairness as aggregation followed by debiasing.\"}]","ALGORITHMIC FAIRNESS IN MACHINE LEARNING: METRICS AND ALGORITHMS - Dissertation | PDF"]