摘要
属性约简是粗糙集理论研究中的关键问题之一.文中定义了一种新的属性重要性度量准则,克服了多值偏向性问题,并给出一种新的属性约简算法.该算法以核属性集为初始约简集合,以新的属性重要性度量准则为启发信息,通过逐步加入相对于决策而言重要的条件属性来求取最小约简.实例分析表明该算法是有效的.
Attribute reduction is one of the key problems in the research on rough set theory. In order to avoid variety bias. a new messure of attribute significance is defined. And based on this method,a new algorithm of attribute reduction is proposed. With the core attributes as the initial reduction,this algorithm uses the attribute significance as heuristic information,and finds the minimal reduction. The resuhs from an example show that this algorithm is effective.
出处
《河南师范大学学报（自然科学版）》
CAS
CSCD
北大核心
2008年第5期163-165,共3页
Journal of Henan Normal University(Natural Science Edition)
基金
安徽省高校省级自然科学研究项目(KJ2008B039)
关键词
粗糙集
属性约简
属性重要性
多值偏向
rough sets
attribute reduction
attribute significance
variety bias
