Research On Customer Churn Risk Early Warning Of The Electricity Business Enterprise In Our Coutry Based On Principal Component And Neural Network

Under the situation of the rapid development on electricity business enterprise in our country,enterprise gradually shift the focus of competition on customer resource competition.With the increase of competitive level on electric business enterprise Customer resources,the customer is more and more in an unstable state,leading to electric business enterprise extended enterprise profit recovery cycle because of the frequent and large amount of customer churn.All the electric business enterprises in such a big data situation with customer churn problem,the electricity companies need to focus on how to predict he future possible customer churn according to the individual characteristics and the individual behavior which customers have reflect,and combined with the behavior rules formulate effective measures,in order to make the enterprise adapt to the change of competitive environment and be closer to the enterprise benefit maximization goal,it is also an important subject in the field of electronic business information management.In order to predict the future customer churn according to customer’s individual and behavior rules,the crucial and the first step is to understand the risk influence factors which will affect enterprise customer churn.For this,I get some important factors which could affect e-commerce enterprises customer churn on the basis of theoretical research combining with interview method and constituting a risk early warning index system of customer churn.Then,in order to access the samples data,investigate the crowd using the target electric business platform in the form of questionnaire.In the research process after questionnaire survey,I applicated science approach to test the reliability and validity of the questionnaire.Then,I applicated the principal component analysis method to extract the most representative indicators,which is most relevant to electricity business enterprise customer churn,as the input layer of the customer loss risk early warning model;use the clustering analysis to get four different customer churn risk early warning levels as the output layer of the model,building the electric business enterprise risk early warning model of customer churn.Finally,after testing the validity of the model,on the basis of studying the article,help the electricity companies to put forward control suggestions to cope with different customer churn risk early warning levels.