The Comparative Analysis and Study of Mobile-Based Customer Data Churn Prediction Model

This paper combined the characteristic of mobile customers churn data, through compare, point out the problem of existing forecast model and the deficiency of traditional vector machine in forecast model. Aimed at the problem of mobile customers churn, established supporting vector machine forecasting model based on ICSVM arithmetic. Through mobile customer data set testing, compare this model to traditional model, this model more adapt to resolve large data set and imbalance data, acquire better customer churn forecasting effect.