基于细分关系的决策表求核与约简算法  

Algorithms of Finding Core and Reduction in the Decision Table Based on the Relation of Subdivision

作　　者：蒋瑜[1] 魏新建[1] 张娟[1] 林和[1] 李永礼[1] 

机构地区：[1]兰州大学信息科学与工程学院,兰州730000

出　　处：《计算机工程与应用》2006年第20期167-169,共3页Computer Engineering and Applications

摘　　要：知识约简是粗糙集理论的核心问题之一。在细分关系下对粗集决策表的核与约简进行了讨论,给出了基于细分关系的求核算法,在此基础上提出了两种决策表约简算法:基于格Hasse图的宽度优先算法和RKCC算法,并结合归纳属性约简算法将这两种算法进行了比较分析。An attribute reduction is a mostly content which the rough sets theory studies.Under the relation of subdivision,this paper analyzes finding core and reduction in the decision of rough set theory,finding core algorithm is developed based on the relation of subdivision,and two attribute reduction algorithms are developed：breadth first algorithm based on the graph of the lattice＇s Hasse and Reduction of Knowledge with Computing Core （RKCC） algorithm.These two algorithms are compared with the attribute inclusion algorithm through theoretical analysis.

关 键 词：核 约简 决策表 细分关系 

分 类 号：TP301[自动化与计算机技术—计算机系统结构]