基于改进差别矩阵的增量式属性约简算法  

Increment algorithm for attribute reduction based on improvement of discernibility matrix

作　　者：冯少荣[1] 张东站[1] 

机构地区：[1]厦门大学信息科学与技术学院,福建厦门361005

出　　处：《深圳大学学报（理工版）》2012年第5期405-411,共7页Journal of Shenzhen University(Science and Engineering)

基　　金：国家自然科学基金资助项目(50604012)~~

摘　　要：研究目前粗糙集中求属性核和属性约简存在的效率低下问题,提出基于改进差别矩阵的核增量式更新算法,用于解决对象动态增加情况下核的更新问题.为降低现有增量式属性约简算法的时间和空间复杂度,提出一种不存储差别矩阵的高效属性约简算法,用于处理对象动态增加情况下属性约简的更新问题.理论及实验结果表明,该算法可明显降低时间和空间的复杂度.An incremental updating algorithm for computing core based on an improved discernibility matrix definition is proposed to improve the efficiency of computing attribute core and attribute reduction in rough sets. This new algorithm is mainly used to solve core updating when objects are dynamically increased. The purpose of this said algorithm is to decrease the complexity of time and space on the existing incremental attribute reduction algo- rithm. The discernibility matrix is not necessarry to be stored and therefore the attribute reduction is updated when objects are dynamically increased. Theoretical analysis and experimental results have shown that this new algorithm is feasible and effective.

关 键 词：粗糙集理论 属性约简 差别矩阵 属性核 决策表 动态更新 增量式算法 知识约简 时间复杂性 空间复杂性 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]