Study And Application Of Data Mining Technology In Zhou Shan Uni-com Churn Prediction

With the development of telecom market accelerating, competition among telecommunication operation companies became more severe. The emphases of the clients relations management is turned from the basic function of the business-accepted, the business, the suit and so on to the control which the customer drains. According to investigates, the cost which the enterprise spends in attracting a new customer is about five to ten times than in maintaining an existing customer. Therefore, the telecommunication operation companies are urgent to figure out how to maintain the existing customer. In order to do it, the telecommunication operation companies need to predict the possibility of the client-draining before clients give up their services. So some corresponding market strategy may be taken. in order to meet the need of the telecommunication operation companies ,The model of client churn in the paper is introduced and provides them a decision support system, which help them find out clients who have high possibility of churn from the massive clients The purpose of this paper is to research and implementation a customer prediction model in telecommunication, it must better accuracy and effectiveness. The research on customer churn prediction of traditional ways focus on using single data mining method to build predictive model, these methods always only give the probability of customer churn, and they don not represent better effect in practice. Based on former research, combining the mobile communications industry status in customer hold management, this paper builds hybrid model for customer churn prediction using decision tree and clusters method. In this paper, a theoretical research and empirical study method is used. Based on China Unicorn ZHOUSHAN branch's customer data, this paper builds hybrid model, including a detailed explanation of the whole process such as attributes choosing, data preparation, construction of the model and model evaluation and application. In this paper, more reasonable evaluation method numerical indicators and the graphic indicators are used to evaluate the result of the model. The result indicates that the hybrid model has better accuracy and hit rates. Meanwhile, the model presents better results than the existing method used by telecom industry at home.