Predicting Telecommunication Customer Churn Using Data Mining Techniques

This paper will illustrate how to use data mining techniques to predict telecommunication customers churn. With a well analysis and interpretation of the data, valuable knowledge and key insights into the customers' needs can be achieved. A sample data based on customer usage was gathered, and different data mining techniques were applied over it. This paper's contribution is to test the capability of a prediction data mining technique, which is the RULES Family algorithm-6 that has never been applied in such a case before. Two pre-stages techniques were applied before the prediction, which are the segmentation "clustering" and the feature selection.