Customer Churn Prediction in Telecommunication Industry. A Data Analysis Techniques Approach

Telecommunications is one of the most dynamic sectors in market, where customer base an important pawn receive safe revenues, so to focus attention paid maintaining them with active status. Migrating customers from network another varies among telecommunication companies depending on different factors such as call quality, pricing plan, minute consumption, data, sms facilities, billing issues, etc. Determining effective predictive model helps detect early warning signals when churn occurs and assigns each a score called “churn score” that indicates likelihood individual might migrate over predefined time period. To this extent, present paper uses more than 10k sample company tries analyse behavior. The aim both test efficiency performance commonly used data mining techniques predict behavior underline main indicators can be conducting analyses. Knowing magnitude phenomenon, prevent instability going occur by applying series measure order increase retention current customers.