[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-122358-en":3,"doc-seo-122358-105":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":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},122358,1374391974468,"Eden","https://ap-avatar.wpscdn.com/davatar_29158cc5080c5b710cf443261637dec0",8,"Research & Report","ADVERSARIAL MACHINE LEARNING FOR SOCIAL GOOD - Dissertation Summary","Adversarial Machine Learning for Social Good examines how machine learning systems used for critical decisions in healthcare, employment, finance, and crime prevention can be made more trustworthy, reliable, and fair. It addresses vulnerabilities to test-time adversarial attacks and explores deception detection generalization to real-life scenarios using multimodal lying features. The work further studies adversarial methods that protect user privacy and security in both vision and tabular loan settings, and proposes multi-concept targeted attacks that reduce one model’s accuracy without harming protected classifiers.","ADVERSARIAL MACHINE LEARNING FOR SOCIAL GOOD  \nby  \nVibha Chandramouli Belavadi  \nAPPROVED BY SUPERVISORY COMMITTEE:  \n\n| Murat Kantarcioglu, Co-Chair |\n| --- |\n| Bhavani Thuraisingham, Co-Chair |\n| Latifur Khan |\n\nRishabh Iyer  \nCopyright © 2022 Vibha Chandramouli Belavadi  \nAll rights reserved  \nDedicated to my family. Mamachi Ajji, Nagu Ajji, Murthy Taata and Rao Taata, this is  \nfor you.  \nADVERSARIAL MACHINE LEARNING FOR SOCIAL GOOD  \nby  \nVIBHA CHANDRAMOULI BELAVADI, BS, MS  \nDISSERTATION  \nPresented to the Faculty of The University of Texas at Dallas in Partial Fulfillment  \nof the Requirements  \nfor the Degree of  \nDOCTOR OF PHILOSOPHY IN  \nCOMPUTER SCIENCE  \nTHE UNIVERSITY OF TEXAS AT DALLAS  \nAugust 2022  \nACKNOWLEDGMENTS  \nI would like to express my appreciation to my advisors, Dr. Murat Kantarcioglu and Dr. Bhavani Thuraisingham. I profoundly thank my advisor, Dr. Murat Kantarcioglu, for his constant guidance to develop strong research foundation, encouragement and teaching me that resilience, patience, and hard work will achieve even the most impossible dream. I would also like to thank my co-advisor, Dr. Bhavani Thuraisingham, for her continuous support and positivity. I would like to extend my most deepest appreciation to Dr. Yan Zhou for her unlimited support and expertise. I am also very grateful to my committee members, Dr. Latifur Khan and Dr. Rishabh Iyer, for their feedback and encouragement. It was a pleasure to work and interact with my labmates Aref Asvadi, Nazmiye Abay, Imrul Anindya, and Mustafa Ozdayi during my PhD.  \nI am most thankful to my parents, Mr. Chandramouli Belavadi and Mrs. Shubha Chandramouli for their unconditional love and belief in me. Thank you for always being my pillar of strength. I also would like to thank my in-laws, Dr. Parampaul Kaur Banipal and Dr. Tarlok Singh Banipal, who have been understanding and supportive during my PhD.  \nTo my husband, Indervir Singh Banipal, thank you for always supporting me and being there for me during the highs and lows of my PhD. journey. This PhD is as much your accomplishment as mine.  \nMay 2022  \nADVERSARIAL MACHINE LEARNING FOR SOCIAL GOOD  \nVibha Chandramouli Belavadi, PhD  \nThe University of Texas at Dallas, 2022  \nSupervising Professors: Murat Kantarcioglu, Co-Chair  \nBhavani Thuraisingham, Co-Chair  \nThe deployment of Machine learning (ML) techniques to automate critical decision-making in healthcare, employment, finance, and crime prevention has played a huge role in improving these systems. However, an incorrect decision can lead to potentially life-changing consequences. In addition, these deployed ML models are also highly vulnerable to test-time adversarial attacks. Thus these ML models need to be made trustworthy and reliable. This dissertation deals with these problems and presents work to make the model more robust, fair, and reliable by using adversarial machine learning techniques.  \nIn this dissertation, we start with the problem of whether ML-based deception detection systems are generalizable to real-life scenarios. We perform experiments to examine whether multimodal aspects such as facial expressions, eye movements, and video cues can be used as deception detection features. We develop three different datasets based on real-life lying scenarios (e.g., for a reward, duress, and speaking white lies) and use them as deception detection features. We also study state-of-the-art deception detection systems and algorithmsand try to extend them to our deception scenarios. We show that deception detection is not generalizable to real-life scenarios, and more subject matter knowledge and better models are needed to make such a claim.  \nWe also address the issue of using adversarial examples to protect user security and privacy and make the black-box models more advantageous to the end-user, specifically in the vision and tabular data domain. In the vision domain, we consider the problems of protecting a sensitive attribute (e.g., ge","cbCainDiixoyWoFf","https://ap.wps.com/l/cbCainDiixoyWoFf","pdf",7503369,1,98,"English","en",105,"# Table of Contents\n## Acknowledgments\n## Abstract\n## List of Figures\n## List of Tables\n## Chapter 1 Introduction\n## Chapter 2 Related Work\n## 2.1 Deception Detection\n## 2.2 Adversarial Attacks\n## 2.3 Multi-concept adversarial examples","[{\"question\":\"Why does the dissertation emphasize making deployed ML models trustworthy and reliable?\",\"answer\":\"Deployed ML models can cause life-changing consequences if decisions are incorrect and are vulnerable to test-time adversarial attacks, so they must be robust, fair, and reliable.\"},{\"question\":\"What does the dissertation find about deception detection in real-life scenarios?\",\"answer\":\"Experiments using multimodal cues and datasets from real lying scenarios show deception detection is not generalizable to real life, indicating the need for more domain knowledge and better models.\"},{\"question\":\"How does the dissertation use adversarial examples to protect privacy and security?\",\"answer\":\"It demonstrates privacy protection by using adversarial artifacts in vision (e.g., obscuring a sensitive attribute) and adversarial recommendations in tabular loan data to steer black-box models toward fairer outcomes.\"}]","ADVERSARIAL MACHINE LEARNING FOR SOCIAL GOOD - Dissertation Summary | PDF",1785810220,247,{"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},"adversarial-machine-learning-for-social-good-dissertation-summary","",{"@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/adversarial-machine-learning-for-social-good-dissertation-summary/122358/",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","2026-08-04",true,{"@type":65,"interactionType":66,"userInteractionCount":20},"InteractionCounter",{"@type":67},"ViewAction",{"@type":69,"mainEntity":70},"FAQPage",[71,77,81],{"name":72,"@type":73,"acceptedAnswer":74},"Why does the dissertation emphasize making deployed ML models trustworthy and reliable?","Question",{"text":75,"@type":76},"Deployed ML models can cause life-changing consequences if decisions are incorrect and are vulnerable to test-time adversarial attacks, so they must be robust, fair, and reliable.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"What does the dissertation find about deception detection in real-life scenarios?",{"text":80,"@type":76},"Experiments using multimodal cues and datasets from real lying scenarios show deception detection is not generalizable to real life, indicating the need for more domain knowledge and better models.",{"name":82,"@type":73,"acceptedAnswer":83},"How does the dissertation use adversarial examples to protect privacy and security?",{"text":84,"@type":76},"It demonstrates privacy protection by using adversarial artifacts in vision (e.g., obscuring a sensitive attribute) and adversarial recommendations in tabular loan data to steer black-box models toward fairer outcomes.","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"]