Churn Prediction In Telecommunications By Data Mining Technology

In mobile telecommunication companies, prepaid customers are not bound by a contract and can therefore change operators at their convenience and without a notification to the company. Loss of a customer, also known as churn customer, to a competitor is a problem facing almost every company in any given industry. This phenomenon is a major source of financial loss, because it is generally much more expensive to attract new customers than it is to retain and sell to existing ones. Therefore, churn is important to manage, especially in industries characterized by strong competition and saturated markets, such as the mobile telecom industry.The main objective of this study is to build a perfect model which is able to predict the customer churning in prepaid system. This research shed light on the issue of customer churn in the telecommunications industry. We used Yemen Mobile as a case study. It is one of companies operating in the mobile telecommunications sector in Yemen suffers from this complicated problem of churn.In this study, we propose and design a model based on the logistic regression as an important and effective tool for classification in data mining. By this technique:· We describe the churning phenomenon, and then give the correct perception and real vision for decision-makers to take precautions and solutions to handle this phenomenon.· We model the problem by defining a set of variables and then apply the logistic regression to find out the coefficient for each variable.· Finally, we present the model as an equation which is able to predict the churn. Applying a various set of statistical tests, we evaluated the accuracy of the proposed model by87.78%.Based on the outcomes of the study, the proposed model can be improved in case that the billing, credit, and demographic data of customers are considered. These data were not available for our study but can provide bases for the interested researchers with good ideas for future researches.