Attributes Reduction for Variable Precision Rough Set Model

XU Hong-yan,E Xu,BAI Xiang-yu,LIANG Xian-zhi,ZHANG Yang

DOI: https://doi.org/10.3969/j.issn.1674-3261.2009.06.005

2009-01-01

Abstract:Attributes reduction was one of the key problems in rough set theory. The attributes reduction of decision table was to simplify the condition attributes of the decision table, after the reduction of the decision table, it kept the same function as that of the former simplication, but it had less condition attributes. Based on the variable precision rough set model, some concepts, such as the reduction of decision table of variable precision rough set model and the core were given their definitions concerned. And a new method for attributes reduction was proposed. This method mainly removed the redundant condition attributes of the main knowledge representation system so as to achieve the reduction of the whole knowledge expression system, and calculated the reduction of the relative core as well. The method not only guaranteed the accuracy of attributes reduction, but also increased its flexibility.