摘要
针对已有的信息集结算子没有考虑属性信息分布疏密程度的情况,研究了一维数据聚类的问题,并给出一种一维聚类的方法.在此基础上,开发了密度加权平均(DWA)中间算子,并将该算子与已有的5种信息集结算子进行合成,得到的合成算子兼顾了多种算子的特点.最后通过一个算例验证了DWA算子及合成算子的一些新颖的特征.
Aiming at the condition that existing information aggregation operators neglect the density degree of attribute information distribution, the theories and a method of one-dimensional clustering are introduced. Based on above theories, the density weighted averaging （DWA） middle operator is developed. Composite operators composed of the DWA and other five existing operators give attention to manifold operators＇ characteristics. Finally, an example shows some novel characteristics of the DWA and composite operators.
出处
《控制与决策》
EI
CSCD
北大核心
2007年第5期515-519,524,共6页
Control and Decision
基金
国家自然科学基金项目(70472032)
关键词
一维聚类
密度加权平均中间算子
多属性决策
One-dimensional clustering
Density weighted averaging middle operator
Multi-attribute decision making
