变精度意义下的优势约简

Dominance reduction on variable accuracy

属性约简是粗糙集理论与方法的研究热点,是一种重要的降维方法.将经典变精度粗糙集模型与优势关系粗糙集模型相结合,定义了优势关系下的β近似集合和β分布协调集的概念,并给出了基于β分布协调集的可辨识属性矩阵定义,提出优势关系下的变精度属性约简方法.通过实例分析,说明了该方法的可行性.

Attribute reduction has been become a research focus in rough set theory. It is a vevy important approach to dimension reduction. This paper proposes a new attribute reduction model,dominance reduction with variable accuracy. The concepts of β approximation sets and β distributed consistent sets are first defined,bastdon dominance relations. Ther,the discernible attribute matrix is constructed by β distributed consistent sets.Finally,an example is illustated to explain the validity of the proposed method.