[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-122483-en":3,"doc-seo-122483-105":30,"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":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},122483,1374391974585,"Genevieve","https://ap-avatar.wpscdn.com/davatar_276721f389ce27ea32af1340a28f341c",8,"Research & Report","BUILDING ROBUST AI SYSTEMS - ADDRESSING UNCERTAINTY, DATA NOISE AND SCARCITY IN MODERN MACHINE LEARNING","Real-world machine learning applications such as autonomous systems, medical diagnostics, and natural language processing require models that remain reliable under uncertainty, data scarcity, and noisy or ambiguous inputs. Conventional methods depending on large clean well-labeled datasets often degrade when labels are limited, inputs are ambiguous, or unlabeled data dominates, producing unreliable predictions and weaker generalization. This thesis proposes robust learning frameworks to improve uncertainty handling, adaptability, reliability, and performance. It introduces HENN for vagueness uncertainty in composite-label classification and proposes NestedMAML for corrupted few-shot learning under noise and out-of-distribution tasks.","BUILDING ROBUST AI SYSTEMS: ADDRESSING UNCERTAINTY, DATA NOISE AND SCARCITY IN MODERN MACHINE LEARNING  \nby  \nChangbin Li  \nAPPROVED BY SUPERVISORY COMMITTEE:  \n\n| Feng Chen, Chair |\n| --- |\n| Rishabh Iyer |\n| Gopal Gupta |\n\nXiaohu Guo  \nCopyright © 2024 Changbin Li All rights reserved  \nTo my wife Jean, my son Aaron, and my parents.  \nBUILDING ROBUST AI SYSTEMS: ADDRESSING UNCERTAINTY, DATA NOISE AND SCARCITY IN MODERN MACHINE LEARNING  \nby  \nCHANGBIN LI, BS, MS  \nDISSERTATION  \nPresented to the Faculty of  \nThe University of Texas at Dallas  \nin Partial Fulfillment  \nof the Requirements  \nfor the Degree of  \nDOCTOR OF PHILOSOPHY IN  \nCOMPUTER SCIENCE  \nTHE UNIVERSITY OF TEXAS AT DALLAS  \nDecember 2024  \nACKNOWLEDGMENTS  \nThis journey was marked by numerous challenges, setbacks, and moments of doubt. Yet, at every turn, I was fortunate to have the unwavering support, guidance, and encouragement of many remarkable individuals.  \nFirst and foremost, I would like to express my deepest gratitude to my PhD advisor, Professor Feng Chen, for his unwavering guidance and mentorship. From the very beginning, he taught me how to approach research from scratch—how to read papers, take notes, debug complex problems, and, most importantly, how to analyze problems and find solutions. His insightful feedback and constant support have been pivotal in shaping both my work and my growth asa researcher.  \nI am also deeply grateful to Professor Rishabh Iyer for his exceptional ability to identify critical issues or blockers in my research. He consistently provided step-by-step suggestions to overcome these challenges, guiding me through complex problems with clarity and precision. His insights and dedication have greatly enriched my understanding and have been instrumental in the progress of this dissertation.  \nI am profoundly grateful to my committee members, Professors Gopal Gupta and Xiaohu Guo. Your insightful feedback and thought-provoking suggestions have significantly shaped and strengthened the direction of my research. Special thanks go to Professors Gopal Gupta and Latifur Khan, whose understanding, support, suggestions, and respect during the tough early stages of my PhD journey were invaluable. Their guidance and encouragement set a strong foundation for my PhD journey. Additionally, I thank Professor Griffith Todd for being the Chair of the Examining Committee.  \nI sincerely thank Dr. Wang Zhou, Dr. Yuxi Hu, Dr. Alex Min, Kaige Liu, Weilin Zhang, Dr. Jason Chen of Meta Platforms, and Dr. Liang Gou, Dr. Sima Behpour of Bosch Center for AI for their mentorship during my internships.  \nI would like to extend my thanks to all my collaborators and lab mates for offering constructive discussions, feedback, and companionship along the way. Special thanks to Dr. Krishnateja Killamsetty, Dr. Suraj Kothawade, and Kangshuo Li, you always provided constructive feedback during our discussion. Especially to Dr. Killamsetty, you always shared your thoughts for new ideas and gave me insights for the research no matter in our project meeting or paper reading. To my exceptional collaborators, Dr. Lance Kaplan, Dr. Audun Jøsang, Dr. Jin-Hee Cho, Dr. Dong Hyun Jeong, Dr. Qi Yu, and Dr. Xujiang Zhao, Kangshuo Li, Yuzhe Ou, Linlin Yu, Tianhao Wang, Rutvij Shah, Haoliang Wang, Kai Jiang, and Raisaat Atifa Rashid, thank you for your dedication and insightful contributions.  \nFinally, to my family, I am forever indebted for their unwavering love and support. A special thanks to my wife, whose understanding, and belief in me have been a constant source of strength throughout this journey. Our relationship has been the greatest blessing of my life, and I feel incredibly fortunate for our marriage and our family, now complete with the arrival of our baby, Aaron. The companionship, empathy, and countless sacrifices you’ve made along the way have left an indelible mark on my heart. I am excited for all the future holds, as we continue to grow together, embrace n","cbCaimLHj0rUTsEj","https://ap.wps.com/l/cbCaimLHj0rUTsEj","pdf",13436832,1,170,"English","en",105,"# Acknowledgments\n## Mentorship and committee support\n## Collaborators and internship guidance\n## Family support\n# Abstract\n## Real-world challenges in machine learning\n## Proposed robust learning frameworks\n## HENN and NestedMAML approaches","[{\"question\":\"What problem does the thesis focus on in modern machine learning?\",\"answer\":\"The thesis focuses on making machine learning models reliable in environments with uncertainty, data scarcity, and noisy or ambiguous inputs.\"},{\"question\":\"Why do traditional machine learning approaches struggle in these settings?\",\"answer\":\"They often rely on large clean well-labeled datasets, so they fail under ambiguous inputs, limited labeled data, or abundant unlabeled data, leading to unreliable predictions and poor generalization.\"},{\"question\":\"What are the two main methods proposed in the thesis?\",\"answer\":\"The thesis introduces Hyper-Evidential Neural Network (HENN) to quantify vagueness uncertainty in composite-label classification and proposes NestedMAML, a nested bi-level optimization framework, to improve robustness in corrupted few-shot learning.\"}]","BUILDING ROBUST AI SYSTEMS - ADDRESSING UNCERTAINTY, DATA NOISE AND SCARCITY IN MODERN MACHINE LEARNING | PDF",1785810894,428,{"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":87,"head_meta":89,"extra_data":91,"updated_unix":28},"building-robust-ai-systems-addressing-uncertainty-data-noise-and-scarcity-in-modern-machine-learning","",{"@graph":36,"@context":86},[37,54,69],{"@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/building-robust-ai-systems-addressing-uncertainty-data-noise-and-scarcity-in-modern-machine-learning/122483/",4,{"url":52,"name":13,"@type":55,"author":56,"headline":13,"publisher":58,"fileFormat":61,"inLanguage":23,"description":14,"dateModified":62,"datePublished":63,"encodingFormat":61,"isAccessibleForFree":64,"interactionStatistic":65},"DigitalDocument",{"name":9,"@type":57},"Person",{"url":41,"name":59,"@type":60},"DocShare","Organization","application/pdf","2026-08-05","2026-08-04",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 thesis focus on in modern machine learning?","Question",{"text":76,"@type":77},"The thesis focuses on making machine learning models reliable in environments with uncertainty, data scarcity, and noisy or ambiguous inputs.","Answer",{"name":79,"@type":74,"acceptedAnswer":80},"Why do traditional machine learning approaches struggle in these settings?",{"text":81,"@type":77},"They often rely on large clean well-labeled datasets, so they fail under ambiguous inputs, limited labeled data, or abundant unlabeled data, leading to unreliable predictions and poor generalization.",{"name":83,"@type":74,"acceptedAnswer":84},"What are the two main methods proposed in the thesis?",{"text":85,"@type":77},"The thesis introduces Hyper-Evidential Neural Network (HENN) to quantify vagueness uncertainty in composite-label classification and proposes NestedMAML, a nested bi-level optimization framework, to improve robustness in corrupted few-shot learning.","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":24},{"code":4,"msg":5,"data":93},[94,98,102,106,111,116,121,124,129,132,136],{"id":20,"doc_module":4,"doc_module_name":46,"category_name":95,"show_sort_weight":96,"slug":97},"Story & Novel",90,"story-novel",{"id":47,"doc_module":4,"doc_module_name":46,"category_name":99,"show_sort_weight":100,"slug":101},"Literature",80,"literature",{"id":53,"doc_module":4,"doc_module_name":46,"category_name":103,"show_sort_weight":104,"slug":105},"Exam",70,"exam",{"id":107,"doc_module":4,"doc_module_name":46,"category_name":108,"show_sort_weight":109,"slug":110},5,"Comic",60,"comic",{"id":112,"doc_module":4,"doc_module_name":46,"category_name":113,"show_sort_weight":114,"slug":115},6,"Technology",50,"technology",{"id":117,"doc_module":4,"doc_module_name":46,"category_name":118,"show_sort_weight":119,"slug":120},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":46,"category_name":12,"show_sort_weight":122,"slug":123},30,"research-report",{"id":125,"doc_module":4,"doc_module_name":46,"category_name":126,"show_sort_weight":127,"slug":128},9,"Religion & Spirituality",20,"religion-spirituality",{"id":127,"doc_module":4,"doc_module_name":46,"category_name":130,"show_sort_weight":127,"slug":131},"World Cup","world-cup",{"id":133,"doc_module":4,"doc_module_name":46,"category_name":134,"show_sort_weight":133,"slug":135},10,"Lifestyle","lifestyle",{"id":137,"doc_module":4,"doc_module_name":46,"category_name":138,"show_sort_weight":107,"slug":139},19,"General","general"]