Customer churn analysis in telecommunication sector

Abstract

Data mining is used to analyze mass databases for having meaningful output. One of the most common applications of the data mining, which is called as Churn Analysis is used to predict behavior of customers who are most likely to change provided service, and to create special marketing tools for them. The aim of this paper is to determine customers who want to churn, and to create specific campaigns to them by using a customer data of a major telecommunication firm in Turkey. To determine the reasons of the customer churn, logistic regression and decision trees analysis, which is one of the classification techniques, are applied.