[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-127238-en":3,"doc-seo-127238-105":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},127238,2336475104042,"Skyler","https://ap-avatar.wpscdn.com/avatar/22000c4c32af1715be0?x-image-process=image/resize,m_fixed,w_180,h_180&k=1786537525561427321",8,"Research & Report","Principled Analysis of Machine Learning Paradigms - Doctoral Dissertation","Deep neural networks used to approximate functions and learn policies for sequential decision-making in high-dimensional complex MDPs have achieved major milestones, yet fundamental understanding of what these models learn and their limiting factors remains incomplete. This thesis introduces reasoning over decisions for analyzing what intelligent agents learn, and proposes a framework to study learned functions, decision-boundary composition, and loss-landscape structure. It provides mathematical reinforcement learning analysis, demonstrates shared high-sensitivity directions across MDPs and algorithms, and establishes geometry-based intrinsic bounds, further contrasting natural versus artificial intelligence.","Principled Analysis of Machine Learning Paradigms  \nEzgi Korkmaz  \nA dissertation submitted in partial fulfillment of the requirements for the degree of  \nDoctor of Philosophy  \nof  \nUniversity College London.  \nDepartment of Computer Science University College London  \nApril, 2024  \n2  \nI, Ezgi Korkmaz, confirm that the work presented in this thesis is my own. Where information has been derived from other sources, I confirm that this has been indicated in the work.  \nAbstract  \nUtilization of deep neural networks as function approximators to learn policies that can make sequential decisions in high-dimensional complex MDPs has led to striking progress for artificial intelligence and machine learning applications. More than a decade of extensive research has enabled us to build artificial intelligence that can transcend natural intelligence in highly complex tasks, yet on the other hand our lack of foundational understanding on the knowledge the artificial intelligences form and the fundamental limitations behind it stands in stark contrast to the exhibited capabilities of these models. In this thesis, first I introduce the concept of reasoning over decisions in artificial intelligence to develop understanding on what truly artificial intelligences learn, and further demonstrate that intelligent agents that reason over their decisions can identify and understand their limitations, capabilities and their knowledge of the world. In the second part of the thesis, I introduce a framework to establish a foundational scientific analysis on the functions learnt by artificial intelligence agents, the underlying composition of the decision boundaries and the structure of the loss landscape. Furthermore, I provide mathematical analysis on the limitations of reinforcement learning that explains the inevitability of high-sensitivity directions in high-dimensions. I conduct extensive empirical analysis in high dimensional complex MDPs and discover that high-sensitivity directions for deep reinforcement learning policies are shared across MDPsand further across algorithms. I further introduce the theoretical foundations establishing that artificial intelligence agents are intrinsically bounded by the geometry of high-dimensions. Both the theoretical analysis I introduce and  \nAbstract 4  \nthe empirical analysis I conduct establish the principles of learning in highdimensions, and the limitations behind it. In the final part of the thesis, I introduce a contradistinction analysis between natural and artificial intelligence. This juxtaposition demonstrates the intrinsic differences between how humans solve the same given tasks compared to intelligent agents. These intrinsic differences between artificial and natural intelligence provide further insights towards understanding how artificial intelligences perceive the world, and the limitations they have that bound their current capabilities to abstract and generalize to complex uncertain environments.  \nImpact Statement  \nArtificial intelligence has carved its place in the daily lives of humanity, from healthcare to foundation models and from autonomous systems to sequential decision making, assisting humans to complete certain tasks and further making decisions for them. While artificial intelligence agents become more and more apart of society, the decisions they make for humanity do have clear and direct effects on the subjects of their decisions. Yet on the other hand, currently we do not know what precisely artificial intelligence agents learn, or how they will behave in complex high-dimensional environments with the knowledge they have.  \nThe framework I introduce in this thesis provides a scientific analysis on the functions learnt by intelligent agents, and the analysis introduced in this thesis establishes a foundational understanding on the limitations and capabilities of artificial intelligence agents. Both the theoretical and empirical analysis provided in this thesis reveal the st","cbCaihFqj1GXnfWo","https://ap.wps.com/l/cbCaihFqj1GXnfWo","pdf",18111873,1,164,"English","en",105,"# Contents\n## Introduction\n## Background\n### Reinforcement Learning\n### Adversarial Formulations","[{\"question\":\"What does the thesis propose to understand what AI models learn?\",\"answer\":\"It introduces reasoning over decisions and develops a framework for scientific analysis of learned functions, decision boundaries, and the loss landscape structure.\"},{\"question\":\"How does the thesis analyze limitations in reinforcement learning?\",\"answer\":\"It provides mathematical analysis explaining the inevitability of high-sensitivity directions in high-dimensional spaces, tying model behavior to intrinsic geometric constraints.\"},{\"question\":\"What empirical findings are reported about deep reinforcement learning policies?\",\"answer\":\"Extensive experiments in complex high-dimensional MDPs show that high-sensitivity directions are shared across MDPs and even across different algorithms.\"}]","Principled Analysis of Machine Learning Paradigms - Doctoral Dissertation | PDF",1785937687,413,{"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},"principled-analysis-of-machine-learning-paradigms-doctoral-dissertation","",{"@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/principled-analysis-of-machine-learning-paradigms-doctoral-dissertation/127238/",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","2026-08-21","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 the thesis propose to understand what AI models learn?","Question",{"text":76,"@type":77},"It introduces reasoning over decisions and develops a framework for scientific analysis of learned functions, decision boundaries, and the loss landscape structure.","Answer",{"name":79,"@type":74,"acceptedAnswer":80},"How does the thesis analyze limitations in reinforcement learning?",{"text":81,"@type":77},"It provides mathematical analysis explaining the inevitability of high-sensitivity directions in high-dimensional spaces, tying model behavior to intrinsic geometric constraints.",{"name":83,"@type":74,"acceptedAnswer":84},"What empirical findings are reported about deep reinforcement learning policies?",{"text":85,"@type":77},"Extensive experiments in complex high-dimensional MDPs show that high-sensitivity directions are shared across MDPs and even across different algorithms.","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"]