[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-119904-en":3,"doc-seo-119904-105":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},119904,3848291630094,"Emma Wilson","https://eur-avatar.wpscdn.com/davatar_085a072bc5b1113ac321206ff7593b45",8,"Research & Report","Feature Bias in Machine Learning Models - An In-depth Exploration for Software Engineering Tasks","This dissertation investigates feature bias in machine learning models for software engineering tasks, focusing on how biased data characteristics affect prediction quality and model interpretability. It studies the role of commit characteristics in defect prediction performance, evaluates the influence of different characteristics, and examines deep learning effectiveness under few-shot class considerations. The work proposes bias-guided misprediction diagnosis and mispredicted area upweight sampling to improve model behavior, and further analyzes attention bias in transformer-based models through empirical measurement and syntax-guided attention patterns.","UNIVERSITY OF CALIFORNIA,  \nIRVINE  \nFeature Bias in Machine Learning Models: An In-depth Exploration for Software  \nEngineering Tasks  \nDISSERTATION  \nsubmitted in partial satisfaction of the requirements  \nfor the degree of  \nDOCTOR OF PHILOSOPHY  \nin Software Engineering  \nby  \nJiri Gesi  \nDissertation Committee: Iftekhar Ahmed, Chair Sam Malek  \nIan Harris  \n© 2023 Jiri Gesi  \nDEDICATION  \nThis dissertation is dedicated to my wife, Jiahui Li, who has been a constant source of support and encouragement during the challenges of doctoral study and life. I am truly thankful for having you in my life. This work is also dedicated to my parents, Shuijin Hua and Yilatu Wu, who have always loved me unconditionally and whose good examples have taught me to work hard for the things that I aspire to achieve.  \nTABLE OF CONTENTS  \nPage  \nLIST OF FIGURES vi  \nLIST OF TABLES vii  \nLIST OF ALGORITHMS ix  \nACKNOWLEDGMENTS x  \nVITA xi  \nABSTRACT OF THE DISSERTATION xiii  \n1 Introduction 1  \n1.1 Background .................................... 6  \n1.1.1 Data Bias in Software Engineering ................... 6  \n1.1.2 Interpret Machine Learning Models ................... 7  \n1.1.3 Self-attention-based Transformer Model ................. 7  \n1.2 Dissertation Structure .............................. 9  \n2 Feature Bias in Software Engineering Machine Learning Model 11  \n2.1 Introduction .................................... 11  \n2.2 Related Work ................................... 14  \n2.2.1 ML for Defect Prediction ......................... 14  \n2.2.2 Few-shot learning and Siamese Network ................ 18  \n2.3 Methodology ................................... 19  \n2.3.1 Prediction Technique Selection ...................... 19  \n2.3.2 Characteristics Selection ......................... 19  \n2.3.3 Investigating Di↵erence in Characteristics between Correct and Incorrect Prediction .............................. 21  \n2.3.4 Investigating Impact of Characteristics on Prediction ......... 23  \n2.3.5 Improving Defect Prediction Considering Few-shot class ........ 25  \n2.4 Results ....................................... 30  \n2.4.1 RQ1: Do commit characteristics have an impact on defect prediction performance? ............................... 31  \n2.4.2 RQ2: Considering di↵erent commit characteristics, which one a↵ects defect prediction performance the most? ................ 32  \n2.4.3 RQ3: How well can DL techniques predict defects by explicitly considering few-shot classes? ......................... 35  \n2.5 Discussion ..................................... 36  \n2.6 Threats to Validity ................................ 39  \n2.7 Conclusions and Future Works .......................... 40  \n3 Leveraging Feature Bias to Interpret Model Misprediction 42  \n3.1 Introduction .................................... 42  \n3.2 Preliminaries ................................... 46  \n3.3 BGMD: Bias Guided Misprediction Diagnoser ................. 48  \n3.3.1 Data Feature Imbalance ......................... 49  \n3.3.2 Bias Guided Misprediction Diagnoser .................. 50  \n3.3.3 Implementation .............................. 55  \n3.4 MAPS: Mispredicted Area uPweight Sampling ................. 55  \n3.4.1 Overview of the baseline algorithms ................... 55  \n3.4.2 MAPS: Mispredicted Area uPweight Sampling ............. 57  \n3.5 Evaluation ..................................... 59  \n3.5.1 ME rule generation technique comparison ............... 60  \n3.5.2 E↵ectiveness of Mispredicted Area Upweight Sampling ........ 64  \n3.5.3 Impact of Upweight Value on MAPS .................. 66  \n3.6 Discussion ..................................... 69  \n3.6.1 Why BGMD works better? ........................ 69  \n3.6.2 Why MAPS is a good method to fix models? ............. 70  \n3.7 Related Work ................................... 71  \n3.8 Threats to Validity ................................ 72  \n3.9 Conclusion ..................................... 72  \n4 Attention Bias in Transformer-base","cbCaiaVHSnofBPmu","https://ap.wps.com/l/cbCaiaVHSnofBPmu","pdf",6348164,1,137,"English","en",105,"# Introduction\n## Background\n## Dissertation Structure\n# Feature Bias in Software Engineering Machine Learning Model\n## Related Work\n## Methodology\n## Results\n## Discussion\n## Threats to Validity\n## Conclusions and Future Works\n# Leveraging Feature Bias to Interpret Model Misprediction\n## Preliminaries\n## BGMD\n## MAPS\n## Evaluation\n## Discussion\n## Related Work\n## Threats to Validity\n## Conclusion\n# Attention Bias in Transformer-based Models for Software Engineering\n## Background\n## Empirical Analysis for Attention Weights\n## SyntaGuid\n## Evaluation\n## Implications\n## Related Works\n## Threats to Validity\n## Conclusion\n# Conclusion","[{\"question\":\"What is the main focus of this dissertation on feature bias?\",\"answer\":\"It examines how biased data characteristics (feature bias) influence machine learning prediction quality and the interpretability of models in software engineering tasks, especially defect prediction.\"},{\"question\":\"How does the dissertation analyze bias in model mispredictions?\",\"answer\":\"It introduces bias-guided misprediction diagnosis (BGMD) to detect bias-related causes and proposes MAPS (mispredicted area upweight sampling) to adjust training emphasis on mispredicted regions.\"},{\"question\":\"How are attention bias and transformer behavior studied in this work?\",\"answer\":\"The dissertation measures attention weights in transformer-based models, analyzes attention bias outcomes, and develops a syntax pattern attention guiding approach (SyntaGuid) using syntax-guided attention patterns.\"}]","Feature Bias in Machine Learning Models - An In-depth Exploration for Software Engineering Tasks | PDF",1785726920,345,{"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},"feature-bias-in-machine-learning-models-an-in-depth-exploration-for-software-engineering-tasks","",{"@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/feature-bias-in-machine-learning-models-an-in-depth-exploration-for-software-engineering-tasks/119904/",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","2026-08-04","2026-08-03",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 is the main focus of this dissertation on feature bias?","Question",{"text":76,"@type":77},"It examines how biased data characteristics (feature bias) influence machine learning prediction quality and the interpretability of models in software engineering tasks, especially defect prediction.","Answer",{"name":79,"@type":74,"acceptedAnswer":80},"How does the dissertation analyze bias in model mispredictions?",{"text":81,"@type":77},"It introduces bias-guided misprediction diagnosis (BGMD) to detect bias-related causes and proposes MAPS (mispredicted area upweight sampling) to adjust training emphasis on mispredicted regions.",{"name":83,"@type":74,"acceptedAnswer":84},"How are attention bias and transformer behavior studied in this work?",{"text":85,"@type":77},"The dissertation measures attention weights in transformer-based models, analyzes attention bias outcomes, and develops a syntax pattern attention guiding approach (SyntaGuid) using syntax-guided attention patterns.","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"]