基于属性近似度的决策信息系统属性约简算法研究  

Research on Attribute Reduction Algorithm of Decision Information System Based on Attribute Approximation

作　　者：何银川 HE Yin-chuan(Department of Computer Technology,Shantou Culture and Art School,Shantou 515065 China)

机构地区：[1]汕头文化艺术学校计算机技术系,广东汕头515065

出　　处：《自动化技术与应用》2022年第7期40-43,61,共5页Techniques of Automation and Applications

基　　金：广东省高职高专云计算与大数据专业委员会项目资助(GDYJSKT19-01);教育部科技发展中心项目资助(2018B01011)。

摘　　要：借鉴已有文献,本文引入属性近似度来探讨不必要属性在不同属性集中作用,给出了P属性及非P属性关于P的属性重要度度量方法。接下来,一方面,本文给出基于核集的启发式属性约简算法,逐步删除备选非核集属性集中不必要属性,降低工作量。另一方面,针对动态增加条件属性的决策信息系统,本文以原约简集为基础,在原约简集中增加必要属性构造新约简集,并在最后反向删除新约简集中不必要属性,完成动态属性约简和精简约简集的任务。最后,通过实例分析验证了该算法的可行性和有效性。After reading the available literatures,this paper proposes the definition of the approximation degree between attribute sets and discusses the role of unnecessary attributes in different attribute sets.In order to measure the significance of attribute,this paper puts forward two different methods to measure the significance of attribute(in or out attribute sets P)about attribute sets P.Then,on the one hand,for the general decision information system,this paper presents a heuristic attribute reduction algorithm based on the kernel set as the starting point,and gradually removes unimportant attributes from nocore attribute sets,which can avoid repeatedly compute those unimportant attributes’significance and reduce its workloads.On the other hand,for the decision information system that dynamically adds conditional attributes,this paper adds necessary attributes in the original reduction set to construct the new attribute reduction set,and finally deletes unnecessary attributes from the new attribute reduction set.All of those can complete the task of quickly solving attribute reduction and simplifying reduction set.Finally,the feasibility and effectiveness of these algorithms are verified through an example analysis.

关 键 词：粗糙集 决策信息系统 属性近似度 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]