In recent years, the telecommunications industry has deeply developed in China. After the restructuring of telecommunication business, the war of contesting customers will become more and more fiercely. It is important to predict which customers would be loss and improve the accuracy of the prediction. With the development of technology of data mining, its importance is obviously, and the field of its application and research is more and more extensive.This paper introduces the method, which builds the prediction models of customers' loss by using three data mining arithmetic-Naive_Bayes, Neural_Network and Decision_Trees. Moreover, the comparison of capabilities and evaluation between these three models was mentioned by this paper. The main contents include:1. Introduced of the development of telecommunications industry in China and the technology of data mining, analyzed the necessity of the application of data minding for customers' loss.2. Introduced the main prediction data mining arithmetic-Naive_Bayes, Neural_Network and Decision_Trees.3. Described the main procedure in the research in details: the source and arrangement of data, the construction of the models, the comparison of capabilities and evaluation between these models. |