ANN-GA approach of credit scoring for mobile customers

In this paper, an approach to evaluate the credit level of mobile customers has been proposed, in order that this would help solve the fraudulent problems that puzzled most of the mobile corps. The approach is based on the combination of the artificial neural network (ANN) and genetic algorithms (GA), which are two useful methods in the field of artificial intelligence, so that it could overcome the drawbacks of both ANN, such as its deficiency in solving the multi-peak problems, and GA, such as the lack of precision compared with ANN, Besides, in the paper we also use the fuzzy sets theory to depict the real state of the customers. The feasibility of the approach has been demonstrated with the comparison between the state and, the credit score of customers with actual data. The approach could be easily applied in the mobile corp. to solve the contradiction between the satisfying services and the efficient management