Churn Prediction Model & Data Warehouse For Churn Prediction

Abstract

This project focus on predicting the potential customers who have the tendency leave from the telecommunication networks, through a prediction model and a Data warehouse solution, which facilitates the prediction model.

In light of high competition the retention of customers with in the telecommunication industry has been a challenging task for every mobile operator in the industry. The project is carried out in order to facilitate end-to-end solutions for churn prediction in telecommunication industry.

Churn prediction model provides the users the facility forecast the potential subscriber base that are about to leave the subjected operator’s subscriber network and to do various range of analysis on churn umj pon historical information. In achieving above core objectives project has completed to implementation. A series of interviews, document study (Desk research), reading through printed material and sketch drawings enabled us to understand requirements.

As system is implemented now and customer is confidant in the use and benefits of this system, future work could be considered. Recommendations as future work are improving this system in terms of predictions, integrating new systems into the Data warehouse and other internal systems such as Accounting. With these future work completed, this system could work as complete enterprise solution on churn prediction for a telecommunication provider.