TITLE:
Churn Forecast Based on Two-step Classification in Security Industry
AUTHORS:
Ying Li, Ziyang Deng, Qiang Qian, Rui Xu
KEYWORDS:
Customer Segmentation, Churn, Self-Organize Map Neural Network, Decision Tree
JOURNAL NAME:
Intelligent Information Management,
Vol.3 No.4,
July
15,
2011
ABSTRACT: Customer is a determinant factor that decides whether a security company will be alive. As a result, the competition for customers is more and more intense between security companies. In order to avoid profit decrease caused by churn, security companies must find those customers who have the loss risk and make measures to maintain loyal customers. Now it is the question that how to find and analyze those customers. In this paper, a two-step classification method about churn Analysis is proposed and the problem of churn in security is analyzed.