基于分辨矩阵与层次分析法的组合属性权重决策方法  

Method for Determining Combined Attribute Weight Based on Discernibility Matrix and Analytic Hierarchy Process

作　　者：吕康 叶军[1,2] 李天淳 LYU Kang;YE Jun;LI Tianchun(School of Information Engineering,Nanchang Insititute of Technology,Jiangxi Nanchang 330099,China;Jiangxi Province Key Laboratory of Water Information Cooperative Sensing and Intelligent Processing,Jiangxi Nanchang 330000,C'hina)

机构地区：[1]南昌工程学院信息工程学院,江西南昌330099 [2]江西省水信息协同感知与智能处理重点实验室,江西南昌330000

出　　处：《河北师范大学学报（自然科学版）》2022年第2期126-138,共13页Journal of Hebei Normal University：Natural Science

基　　金：国家自然科学基金(61562061);江西省教育厅科技项目(GJJ211920,GJ3170995)。

摘　　要：针对基于粗糙集理论与层次分析法结合的决策方法在某些决策表中会出现非核属性权重为0,导致决策精度下降等问题,以属性分辨对象的次数及占比为依据,给出了一种分辨矩阵的属性重要度定义方法,进而提出了分辨矩阵与层次分析法相结合的属性权重决策方法,其不仅避免了非核属性权重为0,而且突出了核属性分辨对象的能力,客观反映了各属性在决策时所起的作用.实例表明组合权重法得到的决策结果更为合理,提高了决策的准确性.For decision-making methods based on the combination of rough set theory and analytic hierarchy process,the weight of non-core attributes will be zero in some decision tables,which will lead to problems such as a decrease in decision-making accuracy.Based on the number and proportion of attribute discrimination objects,we give a method for defining attribute importance of discernibility matrix.On this basis,an attribute weight decision-making method combining discernibility matrix and analytic hierarchy process is proposed.It not only avoids the weight of non-nuclear attributes from being zero,but also highlights the ability of nuclear attributes to distinguish objects,and objectively reflects the role of each attribute in decision-making.The example shows that the decision result obtained by the combined weight method is more reasonable,and the accuracy of the decision is improved.

关 键 词：层次分析法 粗糙集 分辨矩阵 主观权重 客观权重 组合权重 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]