基于矩阵判定分析法和小波网络的银行信贷信用分类方法  

The Synthetic Credit Classification Method with Matrix Discriminant Analysis and Wavelet Network

作　　者：陈凯风[1] 叶中行[1] 

机构地区：[1]上海交通大学数学系,上海200030

出　　处：《上海交通大学学报》2005年第3期448-452,共5页Journal of Shanghai Jiaotong University

基　　金：国家自然科学基金资助项目(10171066);上海市科委重点资助项目(02DJ14063)

摘　　要：分析了矩阵判定分析法和小波网络在解决银行信贷信用分类问题上各自的优缺点,并在此基础上将这两种方法结合起来解决银行信贷信用分类问题.实证分析表明,该模型比单独使用矩阵判定分析法或小波网络的分类准确率要高.Based on the analysis of the advantages and shortages of the matrix discriminant analysis and wavelet network, a new synthetic method which combines the above two algorithms together for credit scoring was introduced. The empirical test shows that this approach is better than the individual of the two methods.

关 键 词：矩阵判定分析法 小波网络 信用评估 

分 类 号：O234[理学—运筹学与控制论]