Research On Personal Credit Evaluation Method Based On Statistical Theory

Since entering the 21 st century,Chinese market economy to develop rapidly,with the improvement of people's material life level,consumption is also growing.A lot of people who didn't have enough money buy commodity,begin to choose to the bank,the loans or installment of the popular way to trade,makes the individual credit business increases quickly,the quantity and amount for the increasing of the personal credit business aeriform in increasing the production of the credit risk,Banks are also facing greater risks as they reap huge profits from their credit operations.Investors in order to reduce the selected high-quality customer,reduce the risk of the investment,forcing them to the credit agencies and industry experts for help,the customer credit assessment.With the increase of the credit business,make the center of the credit agencies and industry experts will work on seeking better personal credit assessment method.Effectively reduce the credit risk,reduce personal mistakes brought about by the economic loss is a problem to be solved.This paper mainly classified prediction of personal credit and personal credit hierarchies are discussed in this paper.First of all,introduce the social significance of the study of personal credit evaluation method and theoretical basis,combined with literature expounds what is personal credit,what is a personal credit assessment;Then in view of the domestic and foreign experts and scholars on the research status of personal credit evaluation method to optimize process in detail;Is given in this paper,the personal credit evaluation method of the main research contents and chapters of this paper structure.The establishment of personal credit evaluation index system.Combined with the literature introduces the evaluation index system of personal credit in domestic and foreign,perfect process step by step to build personal credit evaluation index system in this paper.According to whether a default binary classification problem,using support vector machine(SVM)model for the evaluation of personal credit risk analysis,considering the basic situation,the debt paying ability,stability,impact factor,solving the punishment factor,support vector machine(SVM)model is established for the experimental samples binary classification,and discriminant analysis,C5.0 method and compares the predicted classification accuracy of KNN method,citing the misjudgement of sample support vector machine(SVM)model.To not default,score about sample model is established using discriminant analysismethod to get the discriminant score,then use neural network to predict the sample rating score,get the corresponding prediction score,and the neural network prediction score descending order get ordered sample,the reference of China construction bank personal consumption loan customer credit evaluation method,using DPS data processing system software,carries on the optimal partition of ordered sample,will be no breach of personal credit rating system,in turn,the AAA,AA,A,BBB,BB five grades.Finally,this paper summarizes the research on the personal credit evaluation method,and looks forward to the work that needs to be further studied and improved in the evaluation method.