Study On Credit Risk Based On Artificial Neural Network Model For The Credit Cooperatives Of X County

The thesis firstly reviewed the historical development of the rural creditcooperatives credit risk management, and focused on and analyzed that the currentcredit risk has a great problem. It pointed out that some of the credit risk theoryfoundation more focused on combinations of credit risk management and credit riskcontrol in recent years. But it has not made a larger breakthrough about the credit riskproblem. The purpose of this study was to establish credit risk assessment model ofthe artificial neural network applying to the self-learning ability of the neural networkand nonlinear mapping ability. After that, the paper constructed a BP neural networkmodel that was applied to credit risk assessment. This article were also discussedfrom the theory of network structure and algorithm as well as the indicators of theinput and other issues. And it examples of the artificial neural network to verify thevalidity of the model. Examples were taken from some accounting informationdisclosure of the X county. Finally, the paper constructed the model of artificialneural network for the credit risk assessment. The objective of constructing the fuzzyneural network was to combinate its self-learning capability with the reasoningstructure of the fuzzy system, so as to improve the interpretability of the imitationresults. In accordance with the characteristics of the credit decision, this paperfocused on the three problems, i.e., network structure, learning arithmetic and inputdata processing. The result showed that the artificial neural network not only achievevery high accuracy rate and has a certain interpretation function.This innovation were mainly reflected in two aspects: the first, according to thereal business situations of the X County Rural Credit Cooperatives, its classified anddescribed the current business risks and pointed out the different risk sources and thebank’s influence of its own development. At the same time, according to the reality ofthe X County Rural credit cooperatives credit risk management, its currentdevelopment status were analyzed and discussed. And according to the newnon-performing loans, the deficiencies in current management were pointed out. Thesecond, this paper used the non-performing loan ratio of the X County Rural CreditCooperatives as the research object, using the simplified time series model, and hasmade a tentative empirical study to the evaluation problem of the artificial neuralnetwork model in credit risk. The results of this study showed that using the neural network technology toidentify the credit risk can achieve very high accuracy rate and overcome the credit ofmany uncertain factors and more directly and objectively evaluate it. So it provided ascientific basis to the rural credit cooperatives credit policy and credit riskmanagement. As a complex system powerful tool, the neural network has gooddevelopment prospect in the field of the bank credit risk assessment application.