【作者】 仲秋雁； 莫佳卉；

【Author】 ZHONG Qiu-yan~1 MO Jia-hui~2 (1. Dalian University of Technology, Dalin 116023, China; 2. Harbin Institute of Technology, Harbin 154000, China)

【机构】 大连理工大学管理学院； 哈尔滨工业大学管理学院；

【摘要】 准确的客户市场细分对客户关系管理具有重要作用。现有的聚类算法由于自身的缺陷无法很好的应用在市场细分中。本文借鉴拓展自组织神经网络方法的思想,运用Ward’s指标对SOM聚类结果进行合并,从而得到有效的客群。运用该方法在某零售企业的客户数据中探索出四类客群。ANOVA分析证明了该方法得到结果的可行性,并通过与K—means方法的比较证明了该方法的准确性。更多还原

【Abstract】 Accurate marketing segmentation is important for customer relationship management.Due to the draw- backs of the existing clustering algorithms,they cannot be applied to marketing segments.To get valid customer groups,this research adopted Ward’s index to combine the results of SOM based on the idea of extending SOM. This method was applied to customer samples of one retailing company to achieve four customer groups.ANOVA test demonstrated the results of this method were valid,and the comparison results with K-means proved that this method was accurate.更多还原