一种基于属性集重心的多概念层数据采掘方法

A Method of Mining Multiple-Level Based on Attribute Sets Center of Gravity

文章在属性测度空间和有序分割类概念基础上,仿照力学上的重心概念,定义了属性集的重心及属性重心的高阶中心矩,将属性集重心和多概念层次生成相结合,提出了一种基于属性集重心的多概念层的数据采掘算法,特别是属性集重心概念的引入,极大地方便了人们对属性数据的采掘,提高了数据采掘结果的可信度,符合人们思维推理的习惯。并以年龄概念为例,表明了这种方法直观、切实可行。

In this paper,the concepts of Attribute measure space,ordered cut class and center of attribute sets gravity are proposed.Combining center of attribute sets gravity with conceptual hierarchies,this paper proposes a method of mining Multiple-Level Based on attribute sets center of gravity,especially introduces attribute sets center of gravity.It also makes use of information of original data,and the result of mining data precision is improved.This method is simple and pracitable,and therefore has very important theoretical and practical significance.