[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-118391-en":3,"doc-seo-118391-105":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},118391,1099514067415,"Rowan","https://ap-avatar.wpscdn.com/avatar/100002539d78ffe74a7?x-image-process=image/resize,m_fixed,w_180,h_180&k=1779092875211072502",8,"Research & Report","Machine Learning Explainability Methods - Bachelor’s Thesis","Machine learning models are widely deployed across applications, yet their internal reasoning often remains difficult to interpret. This bachelor’s thesis reviews and evaluates explainability methods with emphasis on tabular-data approaches and neural-network explanations. It compares multiple methods, studies contradictions between explanation outputs, and analyzes how correlated features affect reliability. Based on literature and own experiments, the work proposes conditions or heuristics aiming for consistency among explainable-method results when assumptions are satisfied.","Assignment of bachelor’s thesis  \nTitle: Machine Learning Explainability Methods  \nStudent: Danila Makulov  \nSupervisor: Mgr. Vojtěch Rybář  \nStudy program: Informatics  \nBranch / specialization: Knowledge Engineering  \nDepartment: Department of Applied Mathematics  \nValidity: until the end of summer semester 2023/2024  \nInstructions  \nMachine learning models are increasingly being used in many applications, but they can be difficult to understand and interpret. Explainability is becoming an increasingly important factor in the selection of machine learning models. This thesis aims to review and evaluate various explainability methods for machine learning models.  \n1. Review and summarise the current state of explainability methods for machine learning models. Focus on methods for tabular data and neural networks.  \n2. Evaluate behaviour of at least three explainability methods for models on tabular data(e.g. Individual Conditional Expectations, SHAP, LIME...) and two for neural networks (Saliency Maps, Influential Instances…).  \n3. Try to find examples where various methods contradict each other.  \n4. Analyse the impact of correlated data on explainability results.  \n5. Based on both literature review and own experiments propose conditions or heuristics where explainable methods outputs are consistent.  \nReference:  \n1. Interpretable Machine Learning: A Guide for Making Black Box Models Explainable, [https://christophm.github.io/interpretable-ml-book/](https://christophm.github.io/interpretable-ml-book/) .  \n2. Adebayo, J., Muelly, M., Abelson, H. and Kim, B., 2022. Post hoc explanations may be ineffective for detecting unknown spurious correlation. In International Conference on Learning Representations.  \nElectronically approved by Ing. Magda Friedjungová, Ph.D. on 30 January 2023 in Prague.  \nBachelor’s thesis  \nMACHINE LEARNING EXPLAINABILITY METHODS  \nDanila Makulov  \nFaculty of Information Technology Department of Applied Mathematics Supervisor: Mgr. Vojtˇech Ryb´aˇr January 11, 2024  \nCzech Technical University in Prague Faculty of Information Technology  \n© 2023 Danila Makulov. All rights reserved.  \nThis thesis is school work as defined by Copyright Act of the Czech Republic. It has been submitted at Czech Technical University in Prague, Faculty of Information Technology. The thesis is protected by the  \nCopyright Act and its usage without author’s permission is prohibited (with exceptions defined by the Copyright Act) .  \nCitation of this thesis: Makulov Danila. Machine Learning Explainability Methods. Bachelor’s thesis. Czech Technical University in Prague, Faculty of Information Technology, 2023 .  \nContents  \nAcknowledgments vii  \nDeclaration viii  \nAbstract ix  \nAbbreviation List x  \n1 Introduction 1  \n1. 1 Motivation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1  \n1.2 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1  \n1.3 Aims . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2  \n1.4 Limitations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2  \n1.5 Work Structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2  \n2 Literature Review 3  \n2.1 What is Explainable Machine Learning? ....................... 3  \n2.2 Reasons for xAI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3  \n2.3 What is an explanation?   3  \n2.4 Taxonomy of explainability methods . . . . . . . . . . . . . . . . . . . . . . . . . 4  \n2.5 Interpretable Machine Learning Models . . . . . . . . . . . . . . . . . . . . . . . 4  \n2.5.1 Linear Regression . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4  \n2.5.2 Decision Trees . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5  \n2.6 Global Model-Agnostic Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . 5  \n2.6.1 Partial Dependence Plot . . . . . . . . . . . . . . . ","cbCaikoogkxkHhOE","https://ap.wps.com/l/cbCaikoogkxkHhOE","pdf",4929772,1,73,"English","en",105,"# Introduction\n## Motivation\n## Background\n## Aims\n## Limitations\n## Work Structure\n# Literature Review\n## What is Explainable Machine Learning?\n## Reasons for xAI\n## What is an explanation?\n## Taxonomy of explainability methods\n## Interpretable Machine Learning Models\n## Global Model-Agnostic Methods\n## Local Model-Agnostic Methods\n## The Disagreement Problem\n## Impact of correlation on explainability\n## Neural networks methods\n# Practical Part\n## The Disagreement Problem\n## Brief overview of SHAP and LIME methods\n## Goals of the experiments\n## How often do explanations differ?\n## SHAP vs LIME","[{\"question\":\"What is the main goal of this thesis?\",\"answer\":\"The thesis reviews and evaluates explainability methods for machine learning models, focusing on methods for tabular data and neural networks. It also seeks practical guidance on when explanations are consistent.\"},{\"question\":\"Which explanation methods are evaluated for tabular data and neural networks?\",\"answer\":\"For tabular data, the thesis evaluates at least three methods such as Individual Conditional Expectations, SHAP, and LIME. For neural networks, it evaluates methods including Saliency Maps and Influential Instances.\"},{\"question\":\"How does the thesis address contradictions and correlations in explanations?\",\"answer\":\"It searches for examples where different explainability methods contradict each other. It also analyzes the impact of correlated data on the resulting explanation outputs.\"}]","Machine Learning Explainability Methods - Bachelor’s Thesis | PDF",1785683407,184,{"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},"machine-learning-explainability-methods-bachelors-thesis","",{"@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/machine-learning-explainability-methods-bachelors-thesis/118391/",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","2026-08-02",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 goal of this thesis?","Question",{"text":76,"@type":77},"The thesis reviews and evaluates explainability methods for machine learning models, focusing on methods for tabular data and neural networks. It also seeks practical guidance on when explanations are consistent.","Answer",{"name":79,"@type":74,"acceptedAnswer":80},"Which explanation methods are evaluated for tabular data and neural networks?",{"text":81,"@type":77},"For tabular data, the thesis evaluates at least three methods such as Individual Conditional Expectations, SHAP, and LIME. For neural networks, it evaluates methods including Saliency Maps and Influential Instances.",{"name":83,"@type":74,"acceptedAnswer":84},"How does the thesis address contradictions and correlations in explanations?",{"text":85,"@type":77},"It searches for examples where different explainability methods contradict each other. It also analyzes the impact of correlated data on the resulting explanation outputs.","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"]