Evolutionary churn prediction in mobile networks using hybrid learning

Abstract

Churn is the movement of customers from one mobile network operator to another. It is always better to retain a customer than having to find a new customer in the present competitive environment and the importance of this fact can't be stressed enough. Being more of a social phenomenon than a mathematical one the existing models fail in prediction of such a behavioral quantity. Churn prediction is valuable to the mobile operator depending on the level of accuracy of predictions. This paper presents predictive modeling of customer behavior based on the application of hybrid learning approaches for churn prediction in the mobile network. Our proposed framework deals with a better and more accurate churner prediction technique compared to the existing ones as it incorporates hybrid learning method which is a combination of tree induction system and genetic programming to derive the rules for classification based on the customer behavior. Finally using the game theory techniques we understand the community effect of churn. We calculated the predicted score which is a churn value of a mobile customer. The proposed model is used for prediction of various user defined groupings based on usage time, location and their underlying social network, thus making it a pragmatic approach which models churn on human level than a mathematical level. The post evaluation results on a real world dataset from a leading operator validate our findings. Copyright © 2011, Association for the Advancement of Artificial Intelligence. All rights reserved.