山西大学学报（自然科学版）2025，Vol.48Issue(4)：713-724,12.DOI:10.13451/j.sxu.ns.2024028
对象导出三支概念格的矩阵粗糙熵约简
Matrix Rough Entropy-based Reductions of Object-induced Three-way Concept Lattices
摘要
Abstract
Attribute reduction is an important research issue in formal concept analysis,and seeking efficient attribute reduction meth-ods is also a hot topic of concern.This paper studies attribute reductions of the object-induced three-way concept lattices by introduc-ing matrix rough entropy into the formal context.Firstly,the OE-matrix(object-induced three-way matrix)rough entropy of the OE-concept lattice is defined by combining the OE-object granular matrix with rough entropy,and OE-matrix similarity entropy based on the similarity of OE-object granular matrix is proposed and the related properties of them are discussed.Secondly,the matrix rough entropy consistent set and matrix rough entropy reduction set of the OE-concept lattices are defined.Then the judgment theo-rems for the matrix rough entropy consistent set are given.The importance measure of attributes applying OE-matrix similarity en-tropy is proposed,by which the method and corresponding algorithm for obtaining matrix rough entropy attribute reduction are fur-therly investigated.Finally,experiments of attribute reduction were conducted on the UCI(University of California,Irvine)dataset.The experimental results showed that for datasets with the count of attribute less than 10 and datasets with the count of attribute greater than 10,the reduction time was reduced by an average of 69.1％and 97.5％,respectively,confirming that the algorithm is more suitable for the large sample data.关键词
形式背景/对象粒矩阵/矩阵粗糙熵协调集/矩阵相似熵/属性约简Key words
formal context/object granular matrix/matrix rough entropy consistent set/matrix similarity entropy/attribute reduction分类
信息技术与安全科学引用本文复制引用
胡凯欣,马建敏,刘权芳..对象导出三支概念格的矩阵粗糙熵约简[J].山西大学学报（自然科学版）,2025,48(4):713-724,12.基金项目
国家自然科学基金(61772019) （61772019）
科技部国家重点研发计划项目(2022YFC3303100) （2022YFC3303100）
陕西省自然科学基金(2021JQ-2B) （2021JQ-2B）
