What should lenders be more concerned about? Developing a profit-driven loan default prediction model

作者：

Highlights：

• A novel profit-driven model is proposed to predict loan default.

• Bayesian optimization is used to optimize the hyperparameters of the CBT.

• Profit metric is taken as the optimization objective of the Bayesian optimization.

• SHAP value is calculated to provide interpretable prediction results.

摘要

•A novel profit-driven model is proposed to predict loan default.•Bayesian optimization is used to optimize the hyperparameters of the CBT.•Profit metric is taken as the optimization objective of the Bayesian optimization.•SHAP value is calculated to provide interpretable prediction results.

论文关键词：Loan default risk prediction,Categorical boosting,Bayesian optimization,Profit-driven optimization

论文评审过程：Received 14 June 2022, Revised 15 September 2022, Accepted 27 September 2022, Available online 30 September 2022, Version of Record 14 October 2022.

论文官网地址：https://doi.org/10.1016/j.eswa.2022.118938