Credit scorecards in retail banking: enhancing interpretability through shapley values and evaluating the effectiveness of alternative data for improved accuracy

dc.contributor.advisorRamaboa, Kutlwano
dc.contributor.authorHlongwane, Rivalani
dc.date.accessioned2025-08-26T09:00:37Z
dc.date.available2025-08-26T09:00:37Z
dc.date.issued2025
dc.date.updated2025-08-26T08:57:52Z
dc.description.abstractThis research addresses the dual challenges of improving credit scorecard accuracy and maintaining interpretability. While machine learning algorithms like random forest and eXtreme gradient boosting outperform traditional logistic regression in accuracy, their complex predictor variable representation hinders interpretability. To reconcile this, the study discretizes numerical variables, applies one-hot encoding, and employs Shapley values to derive interpretable credit scores for random forest, eXtreme gradient boosting, light gradient boosting machine, and categorical boosting models. This approach produces credit scorecards that align with industry standards. Additionally, the investigation into the role of alternative data in credit scoring reveals its impact on model accuracy. By analysing unique predictor variables such as an applicant's social circle default status, regional ratings, and local population size, the significance of alternative data is demonstrated. Leveraging the model-X knockoffs framework for predictor variable selection contributes to superior model performance, achieving the highest area under the curve on the Kaggle home credit data.
dc.identifier.apacitationHlongwane, R. (2025). <i>Credit scorecards in retail banking: enhancing interpretability through shapley values and evaluating the effectiveness of alternative data for improved accuracy</i>. (). University of Cape Town ,Faculty of Commerce ,Graduate School of Business (GSB). Retrieved from http://hdl.handle.net/11427/41623en_ZA
dc.identifier.chicagocitationHlongwane, Rivalani. <i>"Credit scorecards in retail banking: enhancing interpretability through shapley values and evaluating the effectiveness of alternative data for improved accuracy."</i> ., University of Cape Town ,Faculty of Commerce ,Graduate School of Business (GSB), 2025. http://hdl.handle.net/11427/41623en_ZA
dc.identifier.citationHlongwane, R. 2025. Credit scorecards in retail banking: enhancing interpretability through shapley values and evaluating the effectiveness of alternative data for improved accuracy. . University of Cape Town ,Faculty of Commerce ,Graduate School of Business (GSB). http://hdl.handle.net/11427/41623en_ZA
dc.identifier.ris TY - Thesis / Dissertation AU - Hlongwane, Rivalani AB - This research addresses the dual challenges of improving credit scorecard accuracy and maintaining interpretability. While machine learning algorithms like random forest and eXtreme gradient boosting outperform traditional logistic regression in accuracy, their complex predictor variable representation hinders interpretability. To reconcile this, the study discretizes numerical variables, applies one-hot encoding, and employs Shapley values to derive interpretable credit scores for random forest, eXtreme gradient boosting, light gradient boosting machine, and categorical boosting models. This approach produces credit scorecards that align with industry standards. Additionally, the investigation into the role of alternative data in credit scoring reveals its impact on model accuracy. By analysing unique predictor variables such as an applicant's social circle default status, regional ratings, and local population size, the significance of alternative data is demonstrated. Leveraging the model-X knockoffs framework for predictor variable selection contributes to superior model performance, achieving the highest area under the curve on the Kaggle home credit data. DA - 2025 DB - OpenUCT DP - University of Cape Town KW - credit scorecard LK - https://open.uct.ac.za PB - University of Cape Town PY - 2025 T1 - Credit scorecards in retail banking: enhancing interpretability through shapley values and evaluating the effectiveness of alternative data for improved accuracy TI - Credit scorecards in retail banking: enhancing interpretability through shapley values and evaluating the effectiveness of alternative data for improved accuracy UR - http://hdl.handle.net/11427/41623 ER - en_ZA
dc.identifier.urihttp://hdl.handle.net/11427/41623
dc.identifier.vancouvercitationHlongwane R. Credit scorecards in retail banking: enhancing interpretability through shapley values and evaluating the effectiveness of alternative data for improved accuracy. []. University of Cape Town ,Faculty of Commerce ,Graduate School of Business (GSB), 2025 [cited yyyy month dd]. Available from: http://hdl.handle.net/11427/41623en_ZA
dc.language.isoen
dc.language.rfc3066eng
dc.publisher.departmentGraduate School of Business (GSB)
dc.publisher.facultyFaculty of Commerce
dc.publisher.institutionUniversity of Cape Town
dc.subjectcredit scorecard
dc.titleCredit scorecards in retail banking: enhancing interpretability through shapley values and evaluating the effectiveness of alternative data for improved accuracy
dc.typeThesis / Dissertation
dc.type.qualificationlevelDoctoral
dc.type.qualificationlevelPhD
Files
Original bundle
Now showing 1 - 1 of 1
Loading...
Thumbnail Image
Name:
thesis_com_2025_hlongwane rivalani.pdf
Size:
1.82 MB
Format:
Adobe Portable Document Format
Description:
License bundle
Now showing 1 - 1 of 1
Loading...
Thumbnail Image
Name:
license.txt
Size:
1.72 KB
Format:
Item-specific license agreed upon to submission
Description:
Collections