Research and Design of Customer Analysis System Based on Clustering Mining

Abstract:

Company customer analysis system is designed and implemented based on clustering mining technology. Customer preference segmentation model is built by using the K-medoids clustering method, the customer are divided into several groups effectively. Through a combination of the decision tree and the logistic regression method, the customer loss early-warning model is constructed. The practical results show that the designed system has an excellent performance.

• Periodical: Applied Mechanics and Materials (Volume 224)
• Pages: 193-198
• DOI: https://doi.org/10.4028/www.scientific.net/AMM.224.193
• Online since: November 2012
• Authors: Jing Lei Xu, Xi Yu Liu
• Keywords: Binary Logistic Regression, Customer Analysis System, Data Mining (DM), Data Warehouse, K-Medoids Clustering

References

[1] Dave Collins, Oracle Essbase & Oracle OLAP, McGraw-Hill/Osborne Media, (2010).

[2] Lv Xiaoling, Xie Bangchang, Data mining method and its application, 1st ed, Beijing: China Renmin University Press, (2009).

[3] YangLin, HuangLiping, On Frame of Synthetic Decision Support System Oriented to CRM , Logistics Technologies, vol. 34(2), 2001, pp.28-30.

[4] Jiawei Han, Micheline Kamber, Data mining Concepts and Technique, 2st edition, Singapore: ElsevierPte Ltd., (2006).

[5] Peng Kai, Qin Yongbin, Xu Daoyun, Customer Stability Modeling Based on Logistic Regression, Computer Enigeerin, vol. 37, 2011, pp.12-15.