企业信用风险评估研究--基于优化的随机森林方法[EB/OL]
北京:中国科技论文在线
构建了企业信用风险评估的随机森林模型
最优分箱WOE变换下的模型分类性能得到了有效提升
通过对输入数据集进行最优分箱和WOE变换
) Abstract: Credit risk is one of the important risks leading to bank failure
How to use the data mining technique to evaluate enterprises' credit risk is of great importance
This paper changed the input data set with the method of optimal binning and WOE transformation, and used the changed data create a new random forest model
To evaluate the model performance, this paper compared the accuracy ratio
如何利用数据挖掘方法对企业信用风险做出准确预测评估具有重大意义
主要研究方向:经济统计 【收录情况】 中国科技论文在线: 汪政元
伍业锋 * ( 暨南大学经济学院
which provides a new vision of interpreting the decision process of random forest
Keywords: credit risk