基于分辨相似矩阵的相似粗糙集的属性约简算法

Attribute Reduction Algorithm Based on Discernable and Similar Matrix About Similar Rough Set

针对现实中数据局限导致等价关系弱化为相似关系,用相似关系代替等价关系建立了相似粗糙集的理论,定义了新的分辨相似矩阵,并给出了基于分辨相似矩阵的求核和属性约简算法,该算法可以有效地减少属性约简的计算量,最后通过实例验证了该算法是适用和有效的。

For weakening equivalent relation to similar relation in information system, the similar rough set by similar relation, the concept ot discernible and similar matrix are defined in the similar information systems. And based on it, the algorithm of attribute reduction and core in similar RS theory is given. Finally, some examples are shown that the minimal reduction of similar information systems can be obtained by using the algorithm simply.