一个新的银行信用风险评估方法

A New Method for Estimating Bank Credit Risk

本文以银行信用风险管理为例,将粗糙集和决策树两种具有互补优势的数据挖掘方法相结合,对客户信用做出归类分析判断,最后利用决策树生成决策规则。实践证明,这种方法忠于原始数据,提高了分类准确度,减小了决策树规模,具有良好的性能。

Stating from the complement between rough new method of data mining based on rough sets and decision sets and decision tree classification algorithm, it proposes a tree classification algorithm, and applies it in the estimating of bank credit risk. Practice has proved that this new method of date mining retains the internal features of the original da- m, speeds up the process of access to knowledge, improves the classification accuracy rate, enhances the interpretability of the rules, and achieves satisfactory results.