[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-117421-en":3,"doc-seo-117421-105":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":4,"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},117421,13056703019404,"Miles","https://ap-avatar.wpscdn.com/davatar_29158cc5080c5b710cf443261637dec0",8,"Research & Report","Predicting Bankruptcy in Industrial Sectors - A Study of Machine Learning Models Performance and Explainability","A master thesis investigates bankruptcy prediction across industrial sectors by comparing classical statistics-based approaches and advanced machine learning models. The study designs a full experimental pipeline: dataset creation and sampling, data quality assessment, preprocessing with scaling, winsorization, and Yeo-Johnson transformation, then model optimization through feature selection and hyperparameter tuning. It evaluates predictive performance on both balanced and highly unbalanced test sets. Finally, it applies explainable AI methods using global and local explanations to interpret model behavior, analyze SHAP-based effects, and discuss results, limitations, and future research directions.","UNIVERSITA’ DEGLI STUDI DI PADOVA  \nDIPARTIMENTO DI SCIENZE ECONOMICHE ED AZIENDALI “M.  \nFANNO”  \nCORSO DI LAUREA MAGISTRALE IN  \nECONOMICS AND FINANCE  \nTESI DI LAUREA  \nPredicting Bankruptcy in Industrial Sectors: A Study of Machine Learning Models Performance and Explainability  \nRELATORE:  \nCH.MO PROF. MICHELE FABRIZI  \nLAUREANDO: SIMONE GISONNA  \nMATRICOLA N. 2089086  \nANNO ACCADEMICO 2024 – 2025  \nContents  \n1 Introduction 5  \n2 Literature Review and Theoretical Background 12  \n2.1 Early Models .................................. 13  \n2.1.1 Risk Index Models ........................... 13  \n2.1.2 Univariate Analysis ........................... 13  \n2.2 Classic statistic-based models ......................... 15  \n2.2.1 Discriminant Analysis ......................... 15  \n2.2.2 Conditional Probability Models .................... 18  \n2.3 Advanced Machine Learning Models ...................... 22  \n2.3.1 Support Vector Machine ........................ 22  \n2.3.2 Tree-based methods ........................... 26  \n2.3.3 Artificial Neural Networks ....................... 32  \n2.4 Explainable AI ................................. 37  \n2.4.1 Global Explanations .......................... 39  \n2.4.2 Local Explanations ........................... 43  \n2.5 Research Questions ............................... 50  \n3 Experimental Environment 51  \n3.1 Dataset and Sampling ............................. 51  \n3.2 Data Quality Assessment ............................ 52  \n3.2.1 Missing Values ............................. 52  \n3.2.2 Class Balance .............................. 55  \n3.3 Data Distribution and Transformation .................... 56  \n3.3.1 Feature Scaling ............................. 58  \n3.3.2 Feature Capping with Winsorization ................. 59  \n3.3.3 Feature Engineering: Yeo-Johnson Transformer ........... 61  \n3.4 Machine Learning Model Optimization .................... 62  \n3.4.1 Feature Selection ............................ 62  \n3.4.2 Hyperarameter Tuning ......................... 64  \n3.4.3 Performances Evaluation ........................ 65  \n3.5 XAI settings ................................... 69  \n3.6 Code Availability and Computational Environment ............. 71  \n4 Results 72  \n4.1 Classification outcomes ............................. 72  \n4.2 Explainability Analysis ............................. 77  \n5 Conclusions 88  \n5.1 Discussion of Key Findings ........................... 88  \n5.2 Limitations and Future Directions ....................... 90  \n5.3 Final Considerations .............................. 92  \nBibliography 92  \nA Supplementary Material 106  \nA.1 Performances .................................. 106  \nA.2 XAI ....................................... 114  \nList of Tables  \n2.1 Definition of Z-Score Components ....................... 16  \n2.2 Median Z-Score by S&P Bond Rating ..................... 18  \n2.3 Set of variables used by Ohlson (1980) .................... 21  \n2.4 Summary of bankruptcy prediction models analysed ............. 36  \n3.1 Final Vector of Variables ............................ 53  \n3.2 Correlation between No. NaNs and target variable .............. 55  \n3.3 Summary of Exploratory Variables ...................... 57  \n3.4 Final Set of Variables .............................. 63  \n3.5 Summary of Models’ Hyperparameters .................... 66  \n4.1 Performance on Balanced Test Set ....................... 72  \n4.2 Performance on 1:100 Unbalanced Test Set .................. 77  \n4.3 Statistical significance of mean SHAP effects ................. 87  \nA.1 Performance with KNN ............................. 108  \nA.2 Performance with SMOTE ........................... 109  \nA.3 Performance with SMOTE and KNN ..................... 110  \nA.4 Performance by industry ............................ 113  \nList of Figures  \n2.1 Graphical example of LDA in a 2 dimensional space (Ning 2023) ...... 16  \n2.2 Hyperplanes separation in R2 , adapted from Burges (1998, p. 129) ..... 23  \n2.3 Non-linea","cbCaielV3qu7aAgk","https://ap.wps.com/l/cbCaielV3qu7aAgk","pdf",4546062,1,118,"English","en",105,"# Introduction\n# Literature Review and Theoretical Background\n## Explainable AI\n# Experimental Environment\n## Dataset and Sampling\n## Data Quality Assessment\n## Machine Learning Model Optimization\n## XAI settings\n# Results\n## Classification outcomes\n## Explainability Analysis\n# Conclusions\n## Limitations and Future Directions","[{\"question\":\"Which machine learning models are evaluated for bankruptcy prediction?\",\"answer\":\"The thesis compares early risk index approaches, classic statistical models, and advanced machine learning models such as Support Vector Machine, tree-based methods, and Artificial Neural Networks.\"},{\"question\":\"How does the thesis handle data preprocessing before training?\",\"answer\":\"It includes missing value assessment, class balance considerations, feature scaling, winsorization (feature capping), and Yeo-Johnson transformation for distributional adjustments.\"},{\"question\":\"What explainability methods are used to interpret model predictions?\",\"answer\":\"The study uses Explainable AI with both global and local explanations, including SHAP-based analyses and other local explanation techniques such as LIME, along with interpretability plots.\"}]","Predicting Bankruptcy in Industrial Sectors - A Study of Machine Learning Models Performance and Explainability | PDF",1785675794,297,{"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},"predicting-bankruptcy-in-industrial-sectors-a-study-of-machine-learning-models-performance-and-explainability","",{"@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/predicting-bankruptcy-in-industrial-sectors-a-study-of-machine-learning-models-performance-and-explainability/117421/",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","2026-08-02",true,{"@type":65,"interactionType":66,"userInteractionCount":4},"InteractionCounter",{"@type":67},"ViewAction",{"@type":69,"mainEntity":70},"FAQPage",[71,77,81],{"name":72,"@type":73,"acceptedAnswer":74},"Which machine learning models are evaluated for bankruptcy prediction?","Question",{"text":75,"@type":76},"The thesis compares early risk index approaches, classic statistical models, and advanced machine learning models such as Support Vector Machine, tree-based methods, and Artificial Neural Networks.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"How does the thesis handle data preprocessing before training?",{"text":80,"@type":76},"It includes missing value assessment, class balance considerations, feature scaling, winsorization (feature capping), and Yeo-Johnson transformation for distributional adjustments.",{"name":82,"@type":73,"acceptedAnswer":83},"What explainability methods are used to interpret model predictions?",{"text":84,"@type":76},"The study uses Explainable AI with both global and local explanations, including SHAP-based analyses and other local explanation techniques such as LIME, along with interpretability plots.","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"]