Establishing An Enterprise Credit Assessment Model Based On Data Mining Techniques

Enterprise credit assessment has become the major risk of commercial banks in China all long. It is a very important but rather complicated process for commercial banks, in which both quantitative data and qualitative data in different aspects must be considered. Making a credit evaluation to bank enterprise customers is an essential way to eliminate the risk for banks. But most credit assessment systems of bank enterprise customers in most researches are over complicated and difficult to apply. The paper introduces a new method "data mining technology" to develop a credit evaluation index system and then establish an enterprise credit assessment model for commercial bank. The data mining method includes principal component analysis, clustering method and Logistic regression method. The rationality of the index system is analyzed by principal component analysis, clustering method and Logistic regression method, and according to which, the index system is reduced to 10 indexes from 22 indexes. At last, the paper constructs an evaluation model by Logistic regression method to predict the enterprise credit risk. Furthermore, the efficiency of the credit assessment model is verified by empirical research.