【摘要】 根据动态交易行为对供应商分类,更好地为供应商提供服务,是大型企业供应商关系管理的核心问题之一。针对供应商行为的交易数据最大、表达复杂的特点,提出基于k-均值子空间聚类算法对供应商分类的数据挖掘方法,解决高维和稀疏数据的分析问题,并通过实例验证该方法的准确性和高效性。结果表明该方法是优化供应商关系,提高企业能力的有效方法。更多还原

【Abstract】 This paper defines that supplier categorization based on dynamic transaction behavior,in order to offer better service for suppliers,is one of core problems in supplier relationship management(SRM).It proposes a K-means type subspace clustering for supplier categorization to analyze the high dimensional and sparse data,because transaction data of supplier behavior is mass and complicated.Therefore,this is a proper way to optimize supplier relations,and improve enterprises′ competence.Finally,a real example is given to demonstrate the effectiveness and veracity of this method.更多还原