摘要
以粗糙集理论为基础 ,引入相似性的概念 ,并提出其衡量方法。改进了粗糙集理论中不可辨关系的确定条件 ,给出了基于新的相似关系的上下近似空间定义 。
Based on rough set theory, the lower and upper approximations are redefined according to a new more general definition using the similarity relation. A general method for defining similarity relations on objects described by a set of attributes is given. An example shows that the decision rules induced from similarity based lower approximations of decision classes are more robust than those induced from indiscernibility based lower approximations.
出处
《控制与决策》
EI
CSCD
北大核心
2001年第5期577-580,共4页
Control and Decision
基金
国家自然科学基金项目 (6993 4 0 2 0 )
