摘要
从粒计算的角度,经典的粗糙集是建立在单一的粒(等价关系)上的,把它推广到建立在优势关系上的多粒度粗糙集,定义了多粒度下的上下近似。通过对经典粗糙集的比较,得到了二粒度和多粒度下粗糙集的一些性质和结论。并在二粒度和多粒度下,对粗糙集里的边界、近似精度、优势度和综合优势度进行了研究。通过地震数据的例子说明了单粒度和多粒度之间的差异。
The classical rough set is based on single granulation（equivalence relation）from the perspective of granu- lar computing, It is extended to the multi-granulation rough set based on dominance relation, and the approximate upper and lower are defined. By comparison of the classical rough set, more properties and conclusions based on two-granulation and multi-granulation are got. And the boundary, approximate accuracy, dominance measure, gener- al dominance measure in rough set are studied. Some examples of seismic data are given to illustrate the differences between single-granulation and multi-granulation.
出处
《计算机工程与应用》
CSCD
2012年第17期53-57,62,共6页
Computer Engineering and Applications
基金
国家自然科学基金(No.10971186
No.11061004)
关键词
优势关系
多粒度
粗糙集
近似精度
优势度
dominance relation
multi-granulation
rough set
approximate accuracy
dominance measure
