Abstract

Background: Large financial companies are perpetually creating and updating customer scoring techniques. From a risk management view, this research for the predictive accuracy of probability is of vital importance than the traditional binary result of classification, i.e., noncredible and credible customers. The customer's default payment in Taiwan is explored for the case study.

Objective: The aim is to audit the comparison between the predictive accuracy of the probability of default with various techniques of statistics and machine learning.

Method: In this paper, nine predictive models are compared from which the results of only six models are taken into consideration. Deep learning-based H2O, XGBoost, logistic regression, gradient boosting, naïve Bayes, logit model, and probit regression comparative analysis are performed. The software tools, such as R and SAS (university edition), are employed for machine learning and statistical model evaluation.

Results: Through the experimental study, we demonstrate that XGBoost performs better than other AI and ML algorithms.

Conclusion: Machine learning approach, such as XGBoost, is effectively used for credit scoring, among other data mining and statistical approaches.

Keywords: Probability, Credit scoring, Artificial intelligence, Statistical model, Deep learning approach, Logistic regression, Gradient boosting.

Graphical Abstract