Based On Neural Network Of Enterprise Crsdit Rating Systems Design And Implementation

Credit risk is the most harmful to the financial markets. Credit risk refers to the possibility of the fiduciary, which is responsible to make debt service that can not perform the expected return and the credit people actually gains deviate. It is the main type of financial risk. Credit risk is directly influence the modern economic life. So, objective comprehensive evaluation of enterprise credit, provide timely and effective basis for decision and avoid credit risk effectively is not only the important problems in the financial sector but also a serious and urgent issue to the academic community. Traditional credit metrics have been unable to meet the needs. In this paper, aiming at the widely used method of enterprise credit evaluation model and the shortcomings, combined with the characteristics of credit assessment, enterprise credit evaluation model was established which based on neural networks. This model includes the determination of score indicator system, the sample data preprocessing and neural network BP algorithm of designs.The Index system of credit rating, which we draw on the existing business indicator system, is based on the characteristics of enterprises in our province. Especially for quantitative indicators, In order to fully reflect the business conditions, it used a wide range of financial indicators and statistical analysis to build the quantitative index system for the enterprise which is selected by the role of large rating, higher ability to identify, less correlation between financial indicators.When obtained the index system in satisfaction, it analyzes the weight of each evaluation in the index system of enterprise credit rating. Then get business credit scoring models. According to the scores of the business credit scoring models, it obtained corporate credit rating of enterprise in the end.The main content of this system include requirements analysis, functional design, interface design, data model design, structural design, and development program options.In this paper, data preparation, the index system model to system modeling, structural design and plan selection all studied carefully one by one. It try to establish a scientific, objective, and advanced corporate credit rating system for enterprise which used by neural network.