Analysis Of Customer Value And Churn Based On Data Mining

In the last few years,with the increasingly fierce competition of the market,customers are grow slowly in many industries.Enterprises are facing more competitive pressures.Customer is the foundation of an enterprise.If there are no new customers for a long time,and with loss of customers,enterprises will be difficult to sustain.So more and more enterprises began to pay attention to the study of customer relationship management.But many research approaches are outdated and subjective,it is not scientific to make judgments about certain attributes in the data.Based on the existing technology,it is necessary to explore the potential information in customer data.It can make rational guidance for enterprise decision.With the development of computer technology,data mining technology emerged in the 1990 s.In recent years,its theoretical knowledge has been improved and the algorithm has been expanded.This paper analyzes customer relationship management based on data mining technology.The analysis of customer relationship management in some articles is monotonous,such as only about customer value,or about customer turnover.But customer value is as important as customer churn,not only focus on how to attract new customers,but also how to retain lost customers and increase customer loyalty.So it is not comprehensive and accurate to research from one side.This article analysis from two perspectives: customer value and customer churn.It used the clustering and classification methods in data mining technology to build a customer value model and a customer churn model,and provide suggestions for the company's marketing strategy based on the model results.This article introduced research background,research status and the framework.Then introduced the Lagrange interpolation and normalization methods used in processing the data and the clustering techniques such as K-Means clustering and AGNES clustering,used in customer value models.This article introduced CART decision tree and LM neural network applied to customer churn model.Then it introduced Logistic regression analysis used in comprehensive analysis.This article take the customer data of an airline company as an example and combined with the theoretical basis,then generate the customer data model.This article analyzed the classification results of the model in detail and given the decision-making advice.Finally,this article explained the aspects of the model to be improved and put forward the outlook.