Research On The Credit Risk Evaluation Of Commercial Banks Based On Bpneural Network

The credit risk is a common problem worldwide, especially in commercial banks.In the West, the theoretical study on credit risk management has formed an effectivesystem, and at the same time the theory has been applied into practice by commercialbanks. But in China, it is not perfect of credit risk management of commercial banks.Due to the theory and techniques are relatively simple, they can not solve all kinds ofproblems. Therefore, it’s significant to put focus on the study of credit risk evaluationof financial institutions,especially commercial banks.In the risk management of commercial banks, credit risk evaluation is the mostimportant part, and it is also the most basic part. Firstly, this paper analyzed thebackground of the research based on reading a large number of domestic and foreignliterature, while a clear focus in this paper should be studied. Secondly, concepts relatedto commercial banks credit risk should be defined, and analysis of the reality of ourcountry commercial banks credit risk evaluation, then introduced the BP neural networkinto the commercial banks credit risk evaluation, and conducted a feasibility analysis.Again, draw lessons from research scholars at home and abroad, combined with thecurrent situation of our country has established the index system of risk evaluationsystem. Including six first-level indexes and eleven second-level indexes, such asenterprise operationability, enterprise profitability and enterprise scale. Finally, collect137financial data of listed companies empirical analysis, factor analysis combined with3δ rule is used to determine the customer’s initial credit level, using Matlab7.0softwarefor the construction of the test model, obtain test results and puts forward relevantpolicy Suggestions.In this paper, through the introduction of the BP neural network, in practice ofcommercial banks credit risk evaluation, using its self-learning ability and adaptiveability to build credit risk evaluation model, then realization the model learning and themodel testing, obtained a high accuracy rate, effectively reduce the artificial factorsinfluence on commercial banks credit risk evaluation, so as to provide certain referencefor commercial banks to guard against credit risk.