A Novel Approach for Churn Prediction Using Deep Learning

Customer retention in telecommunication is one of the prime issue in customer relationship management (CRM). The primary focus of CRM is on existing customer as it is difficult to acquire new customers. The main goal of churn prediction is to classify customers into churner & non-churner. Towards this, deep learning as they are equipped with large increasing data sizes and uncover hidden pattern insights, detects pattern, underlying risks and alert the Telecom Industry about customer behaviour with a better accuracy as compared to the traditional machine learning methods. In this paper, Deep learning by Convolutional Neural Network (CNN) is implemented for churn prediction and it showed good performance in terms of accuracy. The experimental results shows that the predictive model for churn prediction out performs with an accuracy of 86.85%, error rate of 13.15%, precision 91.08, recall 93.08%, F-score 92.06%.