摘要
中国金融市场逐步全面对外资开放,信用卡业务已经成为外资银行争夺的重点,其中导致信用卡风险发生的部分原因是银行系统并没有采取合理的信用评分模型对申请领卡人的资信进行审查。通过将基于统计学习理论的分类方法SVM(Support Vector Machine)引入信用卡申请管理,建立了信用卡申请管理的评分模型。同时将SVM与信用评分领域常用的Logistic回归进行了对比,从而帮助银行挑选优质客户。
With the Chinese banking market gradually opening up to foreign investment, and the credit card business has become the envy of the foreign banks focus. This paper discussed credit card risk which occurred partly because of the bank does not take reasonable credit scoring model for the credit review. Based on the SVM （Support Vector Machine） method, we found an applicant credit scoring model to help selecting high - quality clients. The paper applies the classification method of SVM based on Statistical Learning to the management of credit card applicants. With SVM, a credit scoring model is built for the management of the applicants. Finally, the paper compares the SVM with Logistic Regression which is normally used in the area of credit scoring.
出处
《哈尔滨工业大学学报（社会科学版）》
2007年第4期133-136,共4页
Journal of Harbin Institute of Technology(Social Sciences Edition)
关键词
统计学习
SVM
信用评分
信用卡
statistical learning
SVM
credit scoring
credit card
