优势关系多粒度粗糙模糊集及决策规则获取  

Dominance-based multigranulation rough fuzzy sets and acquisition of decision rules

作　　者：翟永健[1,2] 张宏[1] 

机构地区：[1]南京理工大学计算机科学与技术学院,江苏南京210094 [2]南京交通职业技术学院电子信息工程系,江苏南京211188

出　　处：《江苏大学学报（自然科学版）》2012年第3期310-315,共6页Journal of Jiangsu University：Natural Science Edition

基　　金：国家自然科学基金资助项目(60903027)

摘　　要：为了将多粒度粗糙集方法进一步扩展以适应模糊信息系统的需求,将多粒度思想引入到基于优势关系的粗糙模糊集模型中,提出了基于优势关系的乐观和悲观多粒度粗糙模糊集.在这2种多粒度粗糙模糊集中,采用一族而非一个优势概念来进行目标的逼近,并且被近似的目标是模糊而非清晰的集合.不仅对这2种新的粗糙模糊集的性质进行了讨论,而且研究了如何从模糊信息系统中获取逻辑连接词为"或"的决策规则,并采用一个模糊信息系统对新提出的粗糙集模型及决策规则获取进行了实例分析.结果表明:借助优势关系的方法,可以进一步扩展多粒度粗糙集方法,以处理模糊数据,从而扩大多粒度概念的应用范围.To further generalize the multigranulation rough sets approach into fuzzy information system,the basic idea of multigranulation was introduced into the dominance-based rough fuzzy set to propose dominance-based optimistic and pessimistic multigranulation rough fuzzy sets.In the two multigranulation rough fuzzy sets,a family instead of a single dominance relation was used to approximate fuzzy other than crisp target.Not only the basic properties about the proposed rough fuzzy sets were discussed,but also the problem of ＂OR＂ rules induction was investigated.A fuzzy information system was analyzed to illustrate the new proposed rough set model and acquisition of decision rules.The results show that the multigranulation rough set approach can be further expanded to deal with fuzzy data by dominance relations,and results in expanding the application of multigranulation concept.

关 键 词：粗糙模糊集 多粒度 粗糙集 优势关系 决策规则 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]