High Accuracy Predictive Modelling for Customer Churn Prediction in Telecom Industry

Churn prediction is an important factor to consider for Customer Relationship Management (CRM). In this study, statistical and data mining techniques were used for churn prediction. We use linear (logistic regression) and non-linear techniques of Random Forest and Deep Learning architectures including Deep Neural Network, Deep Belief Networks and Recurrent Neural Networks for prediction. This is the first time that a comparative study of conventional machine learning methods with deep learning techniques have been carried out for churn prediction. It is observed that non-linear models performed best. Such predictive models has the potential to be used in the telecom industry for making better decisions and customer management.