Based On BP Neural Network Of Manufacturing Listed Companies Credit Fisk Assessment Research

Credit risk assessment began in the sixties and seventies of the 20 th century western capitalist countries, led by the United States, the first is the purpose of risk prediction and the situation of the overall operation of the enterprise objective evaluation. Development up to now, the enterprise is an important part of social credit,credit is the foundation to the survival of an enterprise, is very important to the healthy and stable development of market economy. Credit risk assessment as a means of social supervision, preventing risks to its research has a certain practical and theoretical significance.First of all, in this paper, the domestic and foreign research present situation of credit risk assessment to generalize and summarize, defining the credit risk of the enterprise and analysis, and more on the credit risk assessment methods and models commonly used; Secondly, combining with research situation at home and abroad, the applicable manufacturing listed companies in China, summarized the needed for the credit evaluation index, build a set of credit risk assessment index system, index system of the combination of financial and non-financial factors, the 35 indicators are selected, basic covers all the influence factors of enterprise credit assessment, it has certain applicability; Once again, were summarized in this paper, the BP neural network; Finally, randomly selected from the taian database a sample of 198 listed companies manufacturing data, build three layers BP neural network model, and set the reasonable parameter, using the Matlab software to network model of training,make the credit risk assessment index of the output value is close to the target, thus for our country the credit risk evaluation model of listed companies, with not to participate in the training sample data simulation test, compared with the target value,the result confirmed this article built based on neural network of the listed company credit risk assessment model of the assessment result accuracy is 63%, between the particularity of the manufacturing industry of the results of the model error influence,to a certain extent within the scope of the model is still under evaluation results, this paper model of manufacturing listed companies can effectively evaluate credit risk.