一种改进的粗糙集权重方法  

Improved method of attribute weight based on rough sets theory

作　　者：谭宗凤[1] 徐章艳[2] 王帅[2] 

机构地区：[1]广西师范大学数学科学学院,广西桂林541004 [2]广西师范大学计算机科学与信息工程学院,广西桂林541004

出　　处：《计算机工程与应用》2012年第18期115-118,共4页Computer Engineering and Applications

基　　金：国家自然科学基金(No.60963008);广西自然科学基金(No.2011GXNSFA018163)

摘　　要：深入研究粗糙集表征的属性重要度,针对现有粗糙集理论的属性权重确立方法的不足,结合属性集中各个条件属性的重要度及其确定的条件属性取值个数,提出了改进的粗糙集权重的确定方法,分析了其合理性。通过实例说明了,改进的粗糙集权重方法可以解决原有粗糙集权重确定方法存在的问题。This paper deeply studies the attribute importance represented by rough set. Considering the drawbacks of the existing methods of attribute weight assignment based on rough set and both individual significance degree and the number of values of each attribute in the conditional attributes, it proposes an improved method of attribute weight assignment based on rough set, and it＇ s rationality is proved. A case is used to prove that improved weight- ing method can solve the existing problem.

关 键 词：权重 粗糙集 属性重要度 决策表 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]