程度多粒度粗糙集  

Graded Multi-granulation Rough Sets

作　　者：吴志远[1] 钟培华[1] 胡建根[1] 

机构地区：[1]江西农业大学理学院,江西南昌330045

出　　处：《模糊系统与数学》2014年第3期165-172,共8页Fuzzy Systems and Mathematics

基　　金：江西农业大学基金资助项目(09003227);江西农业大学青年基金资助项目(2971)

摘　　要：多粒度粗糙集模型建立在一族而非仅仅一个不可分辨关系的基础上的。在融入一定程度误差的分类思想下,本文在多粒度粗糙集模型基础上将构建程度多粒度粗糙集,其中包括程度多粒度乐观近似算子和程度多粒度悲观近似算子两种形式。讨论了程度多粒度粗糙集的相关性质,并对程度多粒度粗糙集和经典的多粒度粗糙集进行了对比分析,得出了若干具有理论和应用价值的结果,从而为知识获取提供了一个新的不确定性方法。Multi-granulation rough sets is constructed on the basis of a family of the indiscernible relations instead of only one indiscernible relation. The graded multi-granulation rough sets models are proposed based the multi-granulation rough sets models, which include graded multi-granulation optimistic approximation operators and graded multi-granulation pessimistic approximation operators. Not only the properties about the graded multi-granulation rough sets are discussed, but also the relationship between graded multi-granulation rough sets and the classical multi-granulation rough sets are deeply investigated. Some important conclusions in theory and application are achieved. This model gives a new general framework for the study of DM and KDD.

关 键 词：多粒度粗糙集 程度粗糙集 程度多粒度粗糙集 近似算子 

分 类 号：O159[理学—数学]