The development of small and medium-sized enterprises (SaMSE) is related to the country's economic and social development. Most of our enterprises are SaMSE, which play an important role in the employment of urban population, foreign trade and technological innovation. Now, the main source of financing is commercial banks, but commercial banks are also worried about the financing of SMEs. The reason is that, on the one hand, the financial data and other information of SaMSE are not made public; Therefore, it is particularly important to improve the credit rating system for SMEs. The main purpose of this paper is to study the decision-making management model of SaMSE based on ANN. According to the characteristics of artificial neural network (ANN), this paper discusses the applicability and superiority of the ANN model in the credit risk assessment of commercial banks, and proposes a new idea for the development of the credit risk assessment system. The enterprise credit risk assessment model is implemented with matlab tools. Through the example of matlab network training, it is analyzed that the combination of rough set attribute reduction theory and NN method can improve the performance of NN credit evaluation model.