Application of Computational Intelligence to Predict Churn and Non-Churn of Customers in Indian Telecommunication

Abstract

In the modern society, mobile communication became the leading medium of communication. Now the public policies and standardization of mobile communication allows customers to switch from one service provider to another service provider easily. One of the most critical challenges in data and voice telecommunication service industry is retaining customers. The cost of retaining an existing customer is lesser than the cost of getting a new customer. So service providers now shifted their focus from customer acquisition to customer retention. As a result, churn prediction has emerged as the most essential Business Intelligence (BI) application that aims to identify the customers who are about to transfer their service to a competitor i.e. To churn. In this paper, we proposed Counter Propagation Neural Networks (CPNN), Classification and Regression Trees (CART), J48 and fuzzy ARTMAP to predict customer churn and non-churn in telecommunication sector. The dataset analyzed is taken from Indian Telecommunication Service Industry.