Empirical Research about Credit Risk on Neural Network Based Bp Algorithm

Abstract

The paper combines theory with practice and applies neural network technology to establish a credit risk assessment model based on BP neural network technology. The assessment model, to some extent, improves the traditional credit risk analytical approaches in our country, overcoming the defects that subjectivity exists in credit risk measurement, expanding developing route for credit risk measurement and enriching measurement methods in our country's credit risk system. Moreover, model establishment thinking, index selection as well as data choosing are not based on theoretical research, but instead the practical situation during our country's commercial bank loan. As the empirical study indicates that BP neural network credit risk assessment model is effective in commercial banks' credit risk management, and it is feasible in practical application.