摘要
本文在研究变精度粗糙集模型中的近似、分类质量与知识粒度关系的基础上,提出了利用嵌套的等价关系集,可以构造基于不同置信阈值β、不同级别分类质量γ的分层知识粒度,并设计了相应的算法,最后以实例构造了分层知识粒度结构图。
Based on variable precision rough set theory, this paper discusses granularity of knowledge in with approximation and quality of classification, then presents that a nested sequence of equivalence relations can lead to a hierarchical granulation on various levels of confidence β and quality of classification γ. An algorithm is put forward and a figure of hierarchical granulation is given with an example.
出处
《管理工程学报》
CSSCI
2004年第2期60-63,共4页
Journal of Industrial Engineering and Engineering Management
基金
教育部基金资助项目(01JA630048)
关键词
变精度粗糙集
知识粒度
分层知识粒度
variable precision rough set
granularity of knowledge
hierarchical granulation of knowledge
