均值限制相似优势粗糙集及决策分析模型研究  

Mean-limited-similarity-dominance Rough Set Model and Application in Decision Making

作　　者：翟育明[1] 阎瑞霞[2] 李竹宁[1] 郭斌[3] 

机构地区：[1]上海应用技术学院经济与管理学院,上海201418 [2]上海工程技术大学管理学院,上海201620 [3]江西师范大学软件学院,江西南昌330022

出　　处：《数学的实践与认识》2012年第19期127-136,共10页Mathematics in Practice and Theory

基　　金：教育部人文社会科学规划基金(11YJA790165);上海市软科学博士创新基金(11692191400)

摘　　要：针对不完备信息系统中的偏好多属性决策问题,提出了一种基于均值限制相似优势粗糙集的决策分析模型.首先提出了均值限制相似优势关系的概念;然后在均值限制相似优势关系下得到知识的粗糙近似和属性约简,给出了分类决策规则.与相似优势关系和限制相似优势关系比较研究的结果表明:均值限制优势关系的分类精度和质量介于二者之间,而分类误差率则优于相似优势关系和限制相似优势关系,得到的决策规则可信度更高,决策模型与实际情况更加相符.In order to deal with multi-attribute decision-making problems in incomplete information, a mean-limited similarity dominance rough set model is constructed for decision- making problems. Firstly, mean-limited-similarity-dominance relation is presented. Then, a new rough set model based on the relation proposed in this paper is constructed. Knowledge and attribute reduction of this rough set model are studied. Compared similar dominance rela- tion, restrictions similar dominance relation and mean-limited-similarity-dominance relation, we know the classification accuracy and quality of mean-limited-similarity-dominance relation are between the other two. The classification error rate of mean-limited-similarity-dominance relation is better than the other two. The credibility of mean-limited-similarity-dominance relation is high in resulting decision rules. Decision-making model based on a new rough set model proposed in this paper is consistent with the actual situation.

关 键 词：相似优势关系 限制相似优势关系 均值限制相似优势关系 

分 类 号：O225[理学—运筹学与控制论]