Credit risk assessment of bank facility applicant using fuzzy-neural networks (ANFIS)

Financial and credit institutions have always sought to refinement, attract and retain the best investors, advisers, customers, and borrowers in order to be able to maximize their profits from their investments. However, various sciences have tried to provide precise methods for distinguishing their customers. Accordingly, sciences such as psychology, management, mathematics, finance, etc. have sought to achieve this goal. The current study investigates the necessity of using modern data mining techniques in combination with artificial intelligence methods to overcome the complexity of the problem through answering the question of whether the used method would well predict the credit rating of the customers. This, however, the other aspects of the problem, the choice of the most important factors of measurement should not be forgotten. Neural networks are widely used to evaluate credit risk because they have demonstrated their excellent performance for processing nonlinear data with learning capability. However, the problem of neural networks is also in dealing with qualitative information that restricts its applications in practice. To overcome these bugs, in this paper a fuzzy-neural network model is proposed for assessing credit risk. The results of this study indicate that the prediction accuracy of the proposed model is much better than the neural network.

کلیدواژه ها:

Neural network ، Fuzzy system ، Neural-fuzzy network ، Credit risk