The Credit Ranking Research Of Individual Enterprises Based On The Credit Scoring Model

In the national economy, the private enterprises take up a large number and make a great contribution to our social and economic development in the employment,revenue and national fortune. But they confront with many difficult in the self development especially in the financing. The private enterprises need a powerful capital in their development, while the private enterprises usually have a small scale,immature management,low technology and a lower credit rank compared with the national enterprises . the lack of capital will restrict the development of the private enterprises, but in the experience we find many private enterprises have a lower breach of faith than the national enterprises. So it has a very important meaning to build a credit scoring model fitting our private enterprises.The purpose of this research is to establish the suited credit scoring model for the private enterprise. The taken method is to do a quantitative analysis on the base of qualitative analysis, and make the models according to the CRISP-DM process for data mining. In specific, we build Logistic,BP,RBF with Clementine respectively,2 stage mixed model with Logistic-BP,Logistic-RBF, and do a comparative analysis on the forecasting effect for the 5 models. Through the comparative analysis, the final conclusion is as follow: the Logistic-RBF 2 stage mixed model is the optimal model in terms of the forecasting accuracy,the stabilization and the topology structure.