基于关系矩阵的相对粒度增量式属性约简算法  

Relative Granularity Incremental Attribute Reduction Algorithm Based on Relational Matrix

作　　者：汪际和 Wang Jihe(Department of Public and Basic Course Teaching,West Yunnan University of Applied Sciences,Dali,Yunnan 671006,China)

机构地区：[1]滇西应用技术大学公共基础课教学部,云南大理671006

出　　处：《大理大学学报》2024年第12期30-35,共6页Journal of Dali University

基　　金：滇西应用技术大学项目(21JK11,21XN04);云南省教育厅科学研究基金项目(2024J1131)。

摘　　要：属性约简是粗糙集和粒计算理论中一项重要的数据处理方法。针对传统属性约简算法在处理动态数据问题上的不足,在相对粒度的基础上,利用关系矩阵研究了当决策信息系统的对象变化时,相对粒度的增量式变化机制。根据这种变化机制,提出了基于关系矩阵的相对粒度增量式属性约简算法,提高了属性约简效率。Attribute reduction is an important data processing method in rough set and granular computing theory.In view of the shortcomings of traditional attribute reduction algorithms in dealing with dynamic data problems and based on relative granularity,the incremental change mechanism of relative granularity changes is studied using relational matrix when the objects of the decision information system change.According to this change mechanism,a relative granularity incremental attribute reduction algorithm based on relational matrix is proposed to improve the efficiency of attribute reduction.

关 键 词：关系矩阵 相对粒度 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]