基于银行信用卡客户信用的实证研究

In recent years, China's commercial banks vigorously explore the credit card business market, while the promotion and use of credit cards will bring a range of credit risk. This paper attempts to use the support vector machine to study the evaluation of credit card users. Three common kernels are used to establish the SVM model, and we ultimately conclude that the SVM model based on radial basis kernel is optimal. Using this model, bank credit card customers are classified, distinguishing between "good" customers and "bad" customers. This will help banks to take appropriate measures for different people to reduce credit risk.