一个基于神经网络——Logistic回归的混合两阶段个人信用评分模型研究

被引：36

作者：

石庆焱

机构：

[1] 国家统计局统计科学研究所

来源：

统计研究 | 2005年 / 05期

关键词：

信用评分; 神经网络; Logistic回归;

D O I：

10.19343/j.cnki.11-1302/c.2005.05.011

中图分类号：

F224 [经济数学方法];

学科分类号：

0701 ; 070104 ;

摘要：

The paper presents a method to establish a new model to score personal credibility.It firstly establishes a model of neural network.And secondly uses Logistics regression to construct a final model by applying the result (probability of default) of the neural network model.The experimental analysis shows that the new model is more stable than neural network model, and the forecasting accuracy of the new model is higher than simply using Logistic regression method.

引用

页码：45 / 49

页数：5