摘要
正区域是粗糙集理论中的核心概念之一,提高计算正区域算法的效率对于其相关算法的效率有重要的影响。本文提出了一种求正区域的渐增式方法,它能有效地去掉求正区域算法中的一些冗余运算,其计算正区域的时间复杂度为,替换掉该算法的一部分将得到一个属性约简算法,它是高效而往往能满足用户需求的,比较适用于大型数据集。理论结果和实验表明,该方法确实能高效地计算出正区域。
Positive region is one of the basic concepts in rough sets theory . Computing Positive Region effectively is very important for improving the performance of relative algorithms. An incremental method for fast computing positive region is put up with in this paper. This method whose time complexity is can take off some redundant operation. Replacing one part of the method will gain a new algorithm for the reduction of the attributes which is effective and will meet the need of the users. Theoretical analysis and experimental results show that this method can reduce the time complexity effectively,
出处
《微计算机信息》
北大核心
2006年第09Z期266-268,30,共4页
Control & Automation
基金
国家科技成果重点推广项目(No.2003EC000001)资助
关键词
粗糙集
决策表
属性约简
rough set,decision table,attributes deduction