An Application of the Combination of Principal Component Analysis and BP Neural Network to Credit Assessment in the Commercial Banks

摘要

The credit risk assessment is the most difficulty in the credit risk management in commercial banks, so the combination method of principal component analysis and BP neural network based on biology nerve cell was introduced in detail according to the idea of combination forecasting. The principle component analysis is used to handle on input variables in advance to solve the problem of the inefficiency in BP neural network owing to the excessive input variables. Then the combination method is applied to credit risk assessment in the commercial banks through training of the neural network tools and testing. Empirical results and analysis indicate that this kind of method has the satisfied assessment efficiency and theoretical accuracy and is applicable to Chinese commercial banks.