Hierarchical Alpha-cut Fuzzy C-means, Fuzzy ARTMAP and Cox Regression Modelfor Customer Churn Prediction

As customers are the main asset of any organization, customer churn management is becoming a majortask for organizations to retain their valuable customers. In earlier studies, the applicability andefficiency of hierarchical data mining techniques for churn prediction by combining two or moretechniques have been proved to provide better performances than many single techniques over anumber of different domain problems. This paper considers a hierarchical model by combining threedata mining techniques containing two different fuzzy prediction networks and a regression techniquefor churn prediction, namely Alpha-cut Fuzzy C-Means (αFCM), Improved Fuzzy ARTMAP and Coxproportional hazards regression model, respectively. In particular, the first component of thehierarchical model aims to cluster data in two churner and non-churner groups applying the alpha-cutalgorithm and filter out unrepresentative data or outliers. Then, the clustered and representative data asthe outputs are used to assign customers to churner and non-churner groups by the second technique.Finally, the correctly classified data are used to create the Cox proportional hazards model. To evaluatethe performance of the proposed hierarchical model, the Iranian mobile dataset is considered. Theexperimental results show that the proposed model outperforms the single Cox regression baselinemodel in terms of prediction accuracy, Type I and II errors, RMSE, and MAD metrics