Abstract:
There are some objects that deviate the mass in the multiple object ives decision making(MODM) and comprehensive evaluating. So they are called outl ier set. They may affect the result of decision and evaluating if they are rejec ted. It is based on the characters of item set that the algorithms to check outl ier in data mining. And they arent suitable for the outlier checking in MODM and comprehensive evaluating. This article puts forward a clustering algorithm to c heck outlier based on similar coefficient sum and it efficiently solves this pro blem.
摘要: 在多目标决策和综合评价中,有个别对象远远偏离群体,成为孤立点集?如果不 剔除这些对象,就会影响决策和评价的结果?数据挖掘中孤立点集检测算法通常是基于项集 属性的,显然不适于多目标决策(MODM)和综合评价中的孤立点集检测?该文提出了一个基于 相似系数和检测孤立点的聚类算法,有效地解决了这个问题?
关键词:
相似系数,
孤立点,
聚类
CLC Number:
JIANG Lingmin. Clustering Algorithm to Check Outlier Based on Similar Coefficient Sum[J]. , 2003, 29(11): 183-185.
姜灵敏. 基于相似系数和检测孤立点的聚类算法[J]. , 2003, 29(11): 183-185.