均值限制相似优势粗糙集及决策分析模型研究

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.

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

In order to deal with multi-attribute decision-making problems in incomplete information,a mean-limited similarity dominance rough set model is constructed for decisionmaking 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 relation, 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.