A Computational Intelligence based Approach to Telecom Customer Classification for Value Added Services

Customer classification is an imperative task for any organization catering to different market segments. In telecom industry it becomes even more important to identify which value added services(VAS) would be successful with a given customer segment. VAS provide a flexible revenue model that can be customized to different customer segments based on several attributes such as usage and preferences. Selecting and customizing VAS provides a wide canvas to the operators for maximizing their returns on the customer portfolio. Computational intelligence techniques such as Artificial Neural Network (ANN) and Support Vector Machine (SVM) have been successfully used for data mining and machine learning. These techniques provide a mathematical framework for identifying customers profiles and patterns in large datasets that representing the customers' data and their preferences. In this paper, we propose a methodology using SVM and ANN techniques to classify telecom customer data and identify the VAS best suited for the customer segment. We test our results with the SVM yielding high prediction accuracy for the unknown public test data with Radial Basis Function(RBF) Kernel using grid search technique.