Research On Customer Segmentation And Churn Prediction In Telecom Business Analysis System

The telecommunications industry is a high investment,high return and technology and knowledge intensive industries.To gain a competitive advantage,Telecom enterprises must establish long-term relationship with customers.To obtain the maximum profit with the minimum cost by providing high quality service for customers to retain old customers.This can avoid customer churn caused huge losses.There is a large amount of data between enterprises and customers.Through appropriate data mining techniques to help enterprise find the information of decision.So that decision-makers can shorten the time of customer needs and make the right decision at the right time.In this way,telecom enterprises can remain invincible in the increasingly fierce competition in the telecommunications industry.This thesis mainly analyzes the application of data mining in the two themes of customer segmentation and customer churn prediction.According to the specific user transaction data of telecom enterprises,combined with its specific characteristics,firstly,the DB?INDEX criterion is used to optimize the number of clusters,then the K-means algorithm and RStudio software are used to segment the customers,then analysis the customer base after the segment.Based on customer segmentation data of telecom enterprise in particular,propose a method based on local feature points to the space distance of customer churn prediction model combined with RStudio software to train the customer churn prediction.C hoosing the part of the customer data form the customer segmentation results as the test sample set of prediction model to assess the effect of experimental.After a variety of data mining algorithm modeling and multi-dimensional model evaluation,the experimental results show that the prediction accuracy of customer churn prediction model based on point to local feature space is higher than that of other algorithms.The model can improve the accuracy of the prediction of the loss of customers.This can provide some auxiliary for telecom enterprise business decision-making.It is helpful to enhance the competitiveness of telecom enterprises in the communications industry.