摘要
给出一种从连续决策表中提取模糊决策规则的规则提取算法。首先,转化连续属性值为模糊值;然后,给出两个不同对象的模糊属性值关于相应连续属性的相似度;其次,给出了λ相似关系与λ相似类的定义。根据λ相似关系,给出粗糙-模糊空间中的下近似与上近似概念;最后,结合模糊集与粗糙集理论的思想,给出一种从连续值域决策表获取决策规则的算法,并通过实例说明该算法的有效性。
A method of acquiring fuzzy decision rules from continuous domain decision system is proposed. First, we transform the continuous attributes value into the fuzzy value with fuzzy membership functions. Second, we give the definition of similarity degree of two objects of continuous condition attributions. Third, the definitions of λ similarity relation and λ, similarity class of each object are given. According to the 2 similarity relation, we give lower and upper approximations of rough-fuzzy approximation space and their properties. At last, we put forward a method of discovering fuzzy decision rules from continuous domain decision table by combining rough sets and fuzzy sets. Also, we verified the validity of this new algorithm through an example of effectiveness evaluation of the self-repairing flight control system.
出处
《模糊系统与数学》
CSCD
北大核心
2006年第2期127-132,共6页
Fuzzy Systems and Mathematics
基金
ResearchStartFoundationofChangshaUniversityofScienceandTechnology(52129905)
关键词
粗糙集
模糊集
连续属性
λ相似度
λ相似类
Rough Sets
Fuzzy Sets
Continuous Attributes
λ Similarity degree
λ Similarity class