Researches For Credit Risk Rating Of Banks Based On Fuzzy Neural Network

Neural Network and Fuzzy System are simulations of intelligence of the people, Both of them can establish a non-linear model after a given system's input/output data. They have advantages and disadvantages respectively. Because of absorbing the advantages of Neural Network and Fuzzy System, Fuzzy neural network has been widely used in fuzzy control, fuzzy decision-making, expert system, and model identification fields during the past ten years. Therefore, its theory, model, algorithm and application technology have always been a very important question for study in computer field.This paper studies the advantages and disadvantages of Fuzzy Logic System and Neutral Network, and then combines them organically. At the same time these theories are applied to the bank's internal credit rating, which is a great step from theory to practice. Trying to include more rating factors and make the model more scientific, logical and simple, this paper put forward a combined model, which includes BP Neural Network, Fuzzy logic system and Fuzzy neural network. Exerting advantages respectively, this model assesses credit risk based on rating models such as financial risk, basic risk of customer, enterprise's scale in the commercial bank.This paper designed and programmed an assess credit risk rating model on a commercial bank based on MATLAB language system and tool boxes. The assessment result is proved right by our test of new model. Therefore, this new model can be used to assess credit risk in a commercial bank.