Study On Credit Risk Assessment Model Of Commercial Banks Based On Artificial Neural Network

Along with the increasing trend of financial globalization，China graduallyaccelerated the pace of promoting market-oriented interest rate，which leads to theincreasing volatility in the financial world, has brought big challenge to bankersaround the world. In the increasingly competitive living environment, how toscientifically and effectively measure and manage credit risk has a crucial impact onthe commercial bank management. At present, comparing with famous internationalcommercial banks, China’s commercial Banking is still in the development stage, inthe area of credit risk assessment, even more still in a traditional credit analysismethods and the use of rigid rules and regulations, It is difficult to meet thedevelopment needs of the commercial banks. With a strong desire to offer an effectivetechnology methods for our commercial banks, with the application of NeuralNetwork, this paper examines the Commercial bank credit risk assessment from theperspective of commercial banks.On the basis of depth study of commercial bank credit risk assessment modelliterature, this paper introduced the concepts of credit risk and credit risk assessment.Research the main influencing factors of commercial bank credit risk assessmentscientifically by artificial neural network theory, and then set a commercial bank creditrisk assessment index system which contains3levels of27indexes. Improved BPNeural Networks is selected as the commercial bank credit risk assessment model onthe basis of comparison among representative models. Finally, used MATLABsoftware to empirical analysis with144companies’ financial data. The results showthat the discriminate accuracy rate of this model is87.04%, higher than the standardBP neural network and Logistic regression model, which proves that this model couldeffectively assess the credit risk of commercial banks’ corporate customers. Theresults of this research provides a useful method for the commercial bank credit riskassessment, has a certain reference.