Credit Evaluation Based on the Improved Super-Efficiency Method

Abstract

For internet finance companies, in order to differentiate good customers from bad customers, ranking and classifying customers according to credit scores is an important means to avoid risk. In the paper, an improved super-efficiency method is obtained and applied to evaluate the customers' credit in internet finance companies. In contrast to the super-efficiency method, computing results shows the improved method can really overcome two shortcomings of the original method: (1) relative efficiency values of some decision-making units are too large; (2) infeasibility problem can occurred in the original model. Furthermore, compared with other modified super-efficiency method, the improved method is simple. Data analysis illustrates that the improved method can rank and classify customers rationally. So the new method can provide feasible decision basis for risk management.