In this paper, a new approach is presented to identify churn customers in the banking industry. The purpose of this study is to increase the accuracy of the churn customer identification. In order to predict it, the neural network methods of multilayer perceptron, radial basis function, support vector machine, and generalized regression are used. Then, the accuracy is increased by using a NaïveBayes as a meta-classifier. The results show that this approach has led to a significant improvement in the prediction of churn customers. In addition, we figured out that if classifier techniques can achieve good results, the meta-classifier will boost the accuracy considerably.