个人信用评估的一种数据包络分析方法

Study on Chinese personal credit scoring based on data envelopment analysis

对数据包络分析在个人信用评估领域的适用性进行了分析,根据我国的实际情况建立了个人信用评估的指标体系,确定了个人信用评估模型的输入和输出变量并进行了量化处理,并对我国某商业银行个人消费信贷数据进行了实证分析,取得了对于好坏客户的分类准确率能达到89．60%的较为满意的结果。

Througha analyzing many study ies on personal credit scoring we can make a new scoring model of data envelopment analysis （DEA） to settle Chinese personal credit scoring problem;set up a suitable conditions＇ index system of personal credit scoring;confirm the input and the output variables of the personal credit scoring model. Utilize our country some commercial bank personal consumer credit data carrying on a real example analysis,We adopt the method based on structure variables and get a satisfied accuracy rate 89.60% in classifying good and bad customers.