Data mining methods can be divided into four groups according to their functions: prediction model, classification, link analysis and time series prediction. Each function can be developed and modified to adapt to different business applications. For example, classification model can be used to establish credit risk rating model, credit risk rating model, loss model, fraud prediction model and bankruptcy model. In order to realize every function of data mining, there are many different methods or algorithms to use.The credit risk scoring model discussed in this paper mainly belongs to the classification model, so the main methods used are classification analysis and segmentation analysis. The main methods of classification analysis include: decision tree, neural network, difference analysis, logical regression, probability regression; the main methods of segmentation analysis include: K-average, demographic segmentation, neural network segmentation.4.6 factors influencing the score