Research on Bank Credit Risk Assessment Based on BP Neural Network

Credit risk prediction is to forecast whether the borrower can fulfill the financial commitment in advance by taking the relevant information of the customer as the judgment basis. Banks still use traditional mathematical and statistical methods to forecast risks, which cannot meet the growing demand for credit risk forecast business, nor guarantee the quality of forecast. Therefore, this paper studies the application of deep neural network in bank credit risk. This paper presents a bank credit risk forecasting and management model based on multi-level deep neural network. By studying the relationship between each variable and the result of risk prediction, the model further summarizes some features with large impact factors by using the analytic hierarchy process (AHP), and inputs them into the deep neural network, so as to obtain the model’s credit score for customers. The practical application of the system in a commercial bank shows that the system can further improve the accuracy of bank credit risk prediction, improve the management efficiency, reduce the cost of bank credit management, and enhance the competitiveness of the bank.