信息系统中基于可区分对集的属性约简算法研究  

The Research of an Improved Attribute Reduction Algorithm Based on Distinguishable Pair Set in Information System

作　　者：何建仓[1] 侯泽民[1] He Jiancang;Hou Zemin(Department of Information Engineering,Zhengzhou University of Science and Technology,Zhengzhou Henan 450064,China)

机构地区：[1]郑州科技学院信息工程学院,河南郑州450064

出　　处：《信息与电脑》2020年第19期55-57,共3页Information & Computer

基　　金：郑州市科技局自然科学项目(项目编号:20150419)。

摘　　要：本文借鉴已有研究成果,通过认真分析已有属性集可区分能力的度量方法,给出了可区分对集的概念,用于衡量某个属性集的可区分能力。接下来,借助可区分对集,本文给出了相应的属性重要度定义及满足约简的条件,据此设计了一种基于可区分对集的属性约简算法,最后通过实例说明了该算法是有效的。Based on the findings of existing studies,this paper analyzes the existing measures of attribute sets’distinguishable ability,gives the concept of distinguishable pair set which measures attribute sets’distinguishable ability.Then,with the help of distinguishable pair set,this paper proposes the definition of attribute significance and the condition of attribute reduction.After that,this paper designs an improved attribute reduction algorithm based on distinguishable pair set in information system.Finally,this improved algorithm is proved to be effective by an example analyzing.

关 键 词：信息系统 可区分对集 属性约简 

分 类 号：TP181[自动化与计算机技术—控制理论与控制工程]