信息系统中基于属性贴近度的属性约简算法研究  

The research of an improved attribute reduction algorithm based on the close degree of attribute sets in information system

作　　者：何建仓[1] 侯泽民[1] 

机构地区：[1]郑州科技学院信息工程学院,河南郑州450064

出　　处：《科学技术创新》2020年第30期124-126,共3页Scientific and Technological Innovation

基　　金：郑州市科技局自然科学项目(20150419)资助。

摘　　要：借鉴已有属性约简方法,本文从知识划分出发,给出了等价类相似性的度量方法,进而提出了用于衡量属性集之间相似程度的属性贴近度。在此基础上,本文设计了基于属性贴近度的属性约简改进算法,并通过实例分析验证了该算法的有效性,为属性约简提供了新的研究思路。Based on the existing attribute reduction methods,this paper gives the method of how to measure equivalence classes'similarity from the view of the division of knowledge.Then,this paper proposes the concept of the close degree of attribute sets in order to measure the level of similarity between attribute sets.On this basis,this paper designs an improved attribute algorithm based on the close degree of attribute sets.The validity of this algorithm was tested by an example.Finally,this paper provides a new research thought for attribute reduction.

关 键 词：粗糙集 信息系统 属性贴近度 属性约简 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]