基于决策树算法的客户分类模型研究-计算机应用技术专业论文.docxVIP

湖北工业大学硕士学位论文Abstract 湖北工业大学硕士学位论文 Abstract 111e rapid development of information technology accelerates enterprises modernization and informatization．The management information methods represented bv CRM(Customer Relationship Management)make the traditional operation model of enterprises that takes products as focus develop into the advanced management model which is customer oriented． W色11．classifying the customers of enterprises is the premise to carry out CRM effectively．There are many methods for customer classification currently,which have been explained from different angles or by different technologies．It is prevailing that the data mining technology,especially the decision tree algorithm，has been applied in the research of classifying customers．But it hash’t realized the completely effective method for customer classification．This Paper is based on the problem of customer classification in CRM system．focusing on the research of the application of decision tree algorithm in customer classification according to data mining classification technology and the related theories． After analyzing the basic process that the decision tree algorithm iS used to deal with customer classification，this Paper optimizes several aspects of the process and establishes an improved model for customer classification．啊缸S model has improvements respectively in data preprocessing，decision tree generation algorithm and using the rules to classify customers．As the structure of decision tree generated from effective training dataset is relatively simple、析t11 hi曲efficiency,this paper figures out a data preprocessing method based on the theory of 1inear algebra and correlation analysis，which may generate effective training dataset；as the generation rules for decision tree are incomplete．it proposes an expanded decision tree according to the combination of the traditional decision tree and the method based on association rules to mine classification rules，which may generate more compl