《计算机应用研究》|Application Research of Computers

基于最优近似粗糙集的属性约简

Attributes reduction based on optimal approximation set of rough set

免费全文下载 (已被下载 次)  
获取PDF全文
作者 罗来鹏,刘二根,范自柱
机构 华东交通大学 理学院,南昌 330013
统计 摘要被查看 次,已被下载
文章编号 1001-3695(2019)07-004-1940-03
DOI 10.19734/j.issn.1001-3695.2017.12.0856
摘要 为了更好地获取由边界域产生的不确定性规则知识,提出最优近似粗糙集的属性约简方法。给出了近似空间上粗糙集最优近似集的判定与计算,然后引入最优近似分布协调集、最优近似分布约简概念。讨论了Pawlak属性约简、分布约简、最优近似分布约简之间的关系,最后得到在协调决策表中它们是等价的,在不协调决策表中最优近似分布约简是分布约简子集。从UCI数据集选取五个数据集进行实验,结果表明基于最优近似分布约简能够获得较少的约简属性。
关键词 粗糙集; 属性约简; 相似度; 最优近似集; 分布约简; 最优近似分布约简
基金项目 国家自然科学基金资助项目(61472138)
本文URL http://www.arocmag.com/article/01-2019-07-004.html
英文标题 Attributes reduction based on optimal approximation set of rough set
作者英文名 Luo Laipeng, Liu Ergen, Fan Zizhu
机构英文名 School of Sciences,East China Jiaotong University,Nanchang 330013,China
英文摘要 To obtain the knowledge of uncertainty rules generated by the boundary region, this paper developed an attribute reduction algorithm based on the optimal approximate rough set. Firstly this paper presented the judgement and calculation of the optimal approximation set of rough set in approximate space. Secondly, it introduced the concepts of the optimal approximate distribution coordination sets and the optimal approximate distribution reduction, and discussed the relations among Pawlak attributes reduction, distribution reduction and distribution reduction based on the optimal approximation set of rough set. Finally, it obtained the conclusion that they were equivalent in the coordination decision table and the distribution reduction set based on the optipal approximation set of rough set was subset of the distribution reduction sets. Experiment results on datasets from UCI show that the attribute reduction algorithm based on the optimal approximate rough set can achieve less set of attributes.
英文关键词 rough set; attributes reduction; similarity degree; optimal approximation set; distribution reduction; optimal approximation distribution reduction
参考文献 查看稿件参考文献
 
收稿日期 2017/12/30
修回日期 2018/2/27
页码 1940-1942
中图分类号 TP391
文献标志码 A