ABSTRACT
Mobile Internet company each month will produce the loss of customers, especially now market competition become fierce , enterprises to obtain new customers costs rise, mobile internet company operators from their own point of view, customer retention is required for the survival and development of enterprises. A set of data to illustrate the problem very well: the cost of developing a new customer is 4 times to retain an old customer; marketing products to new customers is the success rate for 15%, selling products to existing customers success rate is 50%, to new customers to sell 6 times more expensive than to existing customers to promote sales spend [1]. The goal of this dissertation is to use machine learning classification algorithm, carries on the analysis to the customer over a period of time call behavior and payment behavior and other information, for mobile Internet company find a good customer churn prediction model.
