基于正域向量的决策粗糙集属性约简  

Attribute reduction in decision-theoretic rough set models based on positive region vector

作　　者：黄国顺 HUANG Guoshun(School of Mathematics and Big Data,Foshan University,Foshan 528000,China)

机构地区：[1]佛山科学技术学院数学与大数据学院,广东佛山528000

出　　处：《佛山科学技术学院学报（自然科学版）》2024年第3期14-19,共6页Journal of Foshan University(Natural Science Edition)

基　　金：广东省基础与应用基础研究基金项目(2021B1515120048)。

摘　　要：在决策粗糙集模型中,现有划分层的正域都是通过集合求并所得,但基于该方法的保正域不变的属性约简与基于差别矩阵方法所得约简结果并不一致。提出了一种基于正域向量的决策粗糙集属性约简方法,该方法与基于差别矩阵的约简方法所得结果是一致的。最后给出一个算例说明其一致性。s:In the decision-theoretic rough set model,the existing positive region at the level of classification is integrated by union.However,the attribute reduction based on this method is not consistent with that based on discernibility matrix.This paper analyzes the reason in depth and proposes an attribute reduction based on positive region vector.It is proved that the positive region vector preservation is consistent with the results obtained by discernibility matrix method.Finally,an example is given to illustrate their consistency.

关 键 词：决策粗糙集模型 正域向量 属性约简 差别矩阵 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程] O225[自动化与计算机技术—控制科学与工程]