Research On Heilongjiang Mobile Rural User Pre Leaving Model Based On K-means Algorithm

With the development of the Internet era,the scale of operation and development of China's telecommunications enterprises has been effectively improved.With the innovation of the internal system of telecom enterprises,telecom operators are paying more attention to the changes of the data of telecom customers.Relevant data show that from 2016 to 2017,the customer data of a telecommunication operation enterprise in China was severely off-line with a rate as high as 45%,indicating that the market profitability of the telecommunication operation enterprise is in a negative state.Therefore,in order to change the negative profitability of the market and improve the market competitiveness,the telecom operators need to use the Internet technology and the customer data information stored in the system to mine and analyze the data of customer off-line,so as to ensure the sustainable operation and development mode of the enterprises.At present,for China's telecommunication operators,only by continuously improving the efficiency and quality of telecommunication operation management,can we ensure the low-cost operation of enterprises.Therefore,telecom operators need to use decision tree technology to construct pre-offline customer identification model,and carry out detailed research and explanation from data understanding,data model,result analysis,etc.This can effectively help business managers quickly identify target customer data for a targeted service management.This paper chooses K-means algorithm for clustering analysis and applies it to specific decision-making model.The modeling data comes from the actual user data of Heilongjiang Mobile Company.According to the degree of early warning,we divide the forecast objects into red early warning users,yellow early warning users and blue early warning users.Based on the predicted results of the model,this paper proposes a customer retention scheme of "layered retention and long-term maintenance",including three aspects:(1)Customer stratification(2)Retention measures(3)Establishment of a long-term customer maintenance system.According to the feedback results of Heilongjiang Mobile Company,the forecasting model and retention scheme established in this study have achieved certain results.