基于粗糙集的属性约简算法

Attribute Reduction Algorithm Realization in Rough Set

属性约简是粗糙集理论研究的关键问题之一,现已证明寻找一个决策表的最优约简是N P-hard问题。本文首先介绍可辨识矩阵属性约简的基本算法并对求核算法进行了改进。在此基础上给出决策表中属性重要性的两种度量,并以此为启发式信息,提出了一种属性约简的启发式算法。最后,实验结果表明,该算法在大多数情况下能有效地获得决策表的最优约简。

Attribute reduction is one of the key topics in the rough set theory field. It has been proven that computing the optimal reduction of decision table is an N P - hard problem, Firstly, this paper introduces the basic attribute reduction algorithm in discernibility matrix and the improved core algorithm. Then, based on it two types of significance of attribute in a decision table are defined, then, an algorithm which uses rough set theory with heuristic information is proposed. Finally, the experimental result shows that the algo rithm can obtain the optimal attribute reduction of decision table efficiently in most cases.