一种粗糙集属性约简的改进算法
首发时间:2009-02-20
摘要:属性约简是粗糙集理论的核心内容之一。本文结合粗糙集知识对基于Skowron[1]区分矩阵的属性约简算法进行改进,形成一种简单、有效、普遍适用的属性约简算法。该属性约简算法带着属性的重要度直接约简,是一种适合连续型属性值的冲突决策表的约简方法。经实例计算证明,改算法提高了分类的准确度。
For information in English, please click here
A Modif ied Algorithm of Attribute Reduction in Rough Set
Abstract:Attribute reduction is one of kernel parts researched in Rough Set theory. A modified algorithm of attribute reduction which based on Skowron,discernibility matrix was proposed in this paper, so it is a simple and efficient method for attribute reduction. This method, which reduces with the significance of the attributes directly, suits for the multi continues attributes values of inconsistent decision table. The results show that this way advances the results of categorization.
Keywords: Attribution reduction Rough set Discernibility matrix Decision table
基金:
论文图表:
引用
No.2941339297712351****
同行评议
共计0人参与
勘误表
一种粗糙集属性约简的改进算法
评论
全部评论0/1000