Research On Commercial Banks Credit Rating Based On Neural Network

With the reform and opening up, China has gradually established a modern commercial bank system. In face of rapid modernization and marketization, the commercial bank system are taking large risks, so as the priority among priorities, risk management is at the core position in the process of commercial bank management, and credit risk management occupies the main position. Only by avoiding the bad assets, improving the ratio of outstanding loans, and strictly controlling the credit risks as for as possible can we provide a guarantee for the survival and development of commercial banks.In the era of big data, China’s commercial banks have a leading hardware system and data warehouse, accumulating a large number of data. With the passage of time, the data value will absolutely increase. We should use the historical data to establish an appropriate data mining model to analyze the mass data, extracting the rules and trends buried in the depths of the data, to obtain sufficient information applied to the banking business decisions, provide strong support for the bank’s strategy and business development, to maximize the value of data.This paper make use of the historical loan data, which include financial information and scale information etc., of some commercial bank to establish a model. First of all this paper will quantify all the data, then pre-preprocess the data, and establish the neural network model of credit risk rating with the clean data, calculating the weight of each field in the model, objectively determining the factors that matters. The model show that, the profit margins and return on total assets are important and the Pearson correlation coefficient is very high, so the enterprise’s profit ability is the most important factor influencing enterprises due for repayment of credit. Putting all the enterprises’ information into the neural network model, then we can forecast if the overdue repayment will happen. With the model we can quantitatively analyze the credit risk, which can effectively identify the risk level of the enterprises, helping commercial banks to improve operational efficiency.