模糊决策粗糙集的特定类最小化决策代价属性约简算法  

Attribute Reduction Algorithm for Minimizing Decision Cost of Specific Class in Fuzzy Decision Rough Set

作　　者：卓雪雪 朱苍璐 Zhuo Xuexue;Zhu Canglu(School of Computer Engineering,Anhui Sanlian University,Hefei,Anhui,230000)

机构地区：[1]安徽三联学院计算机工程学院,安徽合肥230000

出　　处：《绥化学院学报》2022年第12期143-149,共7页Journal of Suihua University

基　　金：安徽省“四新”研究与改革实践项目(2021sx046);课程思政示范试点项目(20zlgc018);教育教学改革研究项目(21zlgc061,21zlgc051)。

摘　　要：决策粗糙集模型是粗糙集理论的重要扩展。由于对于实际环境下的信息系统未考虑特定的类,并且数据类型可能是数值型或模糊型,基于此,文章提出一种模糊决策粗糙集的特定类最小化决策代价属性约简算法。文章首先将传统决策粗糙集进行推广,提出了模糊决策粗糙集模型;然后基于特定类的视角,提出了模糊决策粗糙集特定类别的三支决策规则;最后基于最小化决策代价思想,提出了一种特定决策类的最小化代价属性约简算法。实验结果表明,属性约简结果在降低决策代价的同时可以进一步提升信息系统的分类能力。Decision-theoretic rough set model is an important extension of rough set theory.Because the information system in the actual environment does not consider specific classes,and the data type may be numerical or fuzzy,this paper proposes a specific class minimization decision cost attribute reduction algorithm of fuzzy decision-theoretic rough set.In this paper,firstly,the traditional decision-theoretic rough set is extended and the fuzzy decision-theoretic rough set model is proposed;then,based on the perspective of specific class,three-way decision rules of specific class for fuzzy decision-theoretic rough set are proposed;finally,based on the idea of minimizing decision cost,an attribute reduction algorithm of minimizing cost for specific decision class is proposed.The experimental results show that the result of attribute reduction can not only reduce the cost of decision-making,but also further improve the classification ability of information system.

关 键 词：决策粗糙集 特定类 代价矩阵 决策代价 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]