An Empirical Study of Credit Scoring Model for Credit Card

The purpose of this paper is to propose an optimal credit scoring model to reassess the default risk of credit card holders for credit card issuing banks in Taiwan. This paper adopted four credit scoring models which are the linear discriminant analysis, decision tree, back-propagation neural network, and a hybrid method to evaluate the default risk. By comparing the evaluation results of these models, it shows that the decision tree method has the best classification performance in terms of accuracy and sensitivity. These results of this empirical study will be provided to credit card issuing banks for achieving efficient automatically credit reassessment of default risk.