GSM churn management using an adaptive neuro-fuzzy inference system


Karahoca A., Karahoca D., Aydm N.

International Conference on Intelligent Pervasive Computing (IPC 2007), Cheju Isl, South Korea, 11 - 13 October 2007, pp.323-326 identifier identifier

  • Publication Type: Conference Paper / Full Text
  • Doi Number: 10.1109/ipc.2007.119
  • City: Cheju Isl
  • Country: South Korea
  • Page Numbers: pp.323-326
  • Yıldız Technical University Affiliated: No

Abstract

The movement of subscribers from one operator to another operator is named as churn management for looking for better and cheaper products and services. As markets become saturated and competition intensifies, customers have more choices to take promotions from alternative telecom operators in Turkish GSM (Global Services of Mobile Communications) sector. This study compares various data mining techniques to obtain best practical solution for churning customer detection. Test results offer the Adaptive Neuro Fuzzy Inference System (ANFIS) as a means to efficient churn management methodology. The test bed results show that ANFIS provides 85% of sensitivity with 88% of specificity where it classified 80% of the instances correctly.