A Study Of Automobile Customer Portrait And Churn Prediction Based On Data Mining

Since China's accession to the WTO,its automobile industry has undergone rapid development over the golden decade(2002-2012).After the great increase of urban auto ownership,the sales of passenger vehicles have witnessed slow growth since 2013.Meanwhile,affected by market demand,passenger vehicles manufactured by different automobile enterprises share less difference in shape,performance and price,therefore leading to severe homogeneity.Under the business environment of continuous saturation of passenger vehicles ownership and fierce competition with their counterparts in China,precise positioning of customer demand and prevention of customer churn are the focuses of major auto brands.The accumulation of operating data and the continuous development of data mining technology motivate enterprises to focus on their internal data resources.Data mining technology has become an important means of market competition.The present thesis starts from auto customer data of business dealer management system and then applies R language and SPSS 18 to make customer image analysis and customer churn prediction.The main research work includes not only the use of K-means cluster analysis and Apriori correlation analysis to identify customer characteristics and their needs and depict their five distinct characteristics;but also the application of data mining technology on the basis of decision tree to model the problem of customer churn in auto sales and then classify potential loss of customers according to the prediction results of model classification.The results of this paper can directly instruct work like marketing and customer relationship maintenance.First of all,the group characteristics,hobbies and needs of auto buyers obtained by cluster analysis and correlation analysis can provide evidence for sales departments to make differentiated marketing strategies.Secondly,make use of the prediction results of the decision tree classification model to divide VIP customer groups,therefore making targeted care and repair to prevent customer churn.