中文摘要：

聚类在商业研究中广泛应用于对未知特征的客户群进行价值细分。分析了传统细分方法对于细分客户初始条件敏感的弱点，多准则客观聚类方法不需要借助领域专家的知识，能够自动、客观地确定聚类个数及最优聚类方案。本文采用多准则客观聚类方法进行客户价值细分。提出新一致性准则体系及算法的实施步骤，并将其应用于客户价值细分中。通过实证对比研究．结果表明新算法具有无需预先指定聚类数、抗干扰性强的特点，使得细分结果更加准确。

英文摘要：

Clustering is widely used in customer value segmentation in business researches. This paper points out the shortcomings typical of traditional techniques and presents multi-criteria objective clustering analysis, a new method for value-driven customer segmentation, which compared with other clustering methods, can determine the number of clusters and find out the optimal clustering scheme in an automatic and objective way independent of specific expertise. The new criterion is followed and be used in segmentation. The results indicate that MEC-OCA may be more useful for customer segmentation because it provides more homogeneous segmentation solutions than the other clustering algorithm, and is without initial category numbers and noise resistance. The advantage of the new ones is proved through theoretical analysis and demonstration.