The Research On Personal Credit Evaluation In Commercial Banks

At present, China's personal credit system is not yet perfect, the personal credit growing trend of rapid development. Credit Commercial banks accounted for the overall proportion of loans on the rise, has become the bank's expected profit growth. Strengthen the personal credit assessment methods and targets of the study, may greatly reduce the banks to develop consumer credit risk, accelerate the development of personal credit. In this paper, the study of the results of in-depth analysis on the basis of elaborated discriminant analysis, linear regression, Neural network classification tree model such as the basic principles. Then on the domestic and international personal credit assessment of the contents of detailed analysis and comparison Accordingly suitable for the establishment of China's actual situation in personal credit assessment index system. This paper from China's commercial banks actual personal credit business from the appropriate samples, Application classification tree model of the separation operation, the impact of the classification of the model structure of the sample, wrong parameters, such as the cost of different combinations of tests, the classification tree model optimization, has been applied to commercial banks for the personal credit assessment model. Then were used neural network algorithm EXhaustive prune, Prune algorithm, RBFN algorithm modeling, and model optimization. Finally, the two models are compared, and the results of the study showed that Neural Network Model EXhaustive prune algorithm is better than the C5 Decision Tree Algorithm.