摘要
针对改进的经典算法求取属性约简存在的时间和空间性能不理想问题,本文提出了一种新的属性约简算法ARSDM。该算法先将决策表按决策属性的类别划分,后采用边生成矩阵元素边约简边排序的思想求取属性约简,有效地加快了约简速度。实验表明ARSDM算法与经典算法相比具有较好的时间和空间性能。
Traditional algorithm has relatively poor efficiency in both time and space when obtaining attribute reduction. Based on SDM, a new algorithm of attribute reduction called ARSDM is proposed in this paper. ARSDM takes the idea of classifying the universe of decision table according to the value of decision attribute firstly, then reducing and sorting elements of SDM while constructing them. The experimental study shows that the algorithm of ARSDM out- performs the traditional algorithms largely on both time and space.
出处
《计算机科学》
CSCD
北大核心
2008年第3期209-212,共4页
Computer Science
基金
安徽省自然科学基金(050420207)
关键词
数据挖掘
粗糙集
不一致性决策表
属性约简
分辨矩阵
Data mining, Rough set, Inconsistent decision table, Attribute reduction, Discernability matrix