摘要
本文在[1]的基础上研究了一类单值选择函数的和型Fuzzy实现的有关问题,首次将此类问题转化为神经元理论中的感知器模型。本文在有限社会状态集合的情形,给出了单值选择函数是和型Fuzzy可实现的一个充要条件;并给出了相关几何描述结论;借助于感知器的学习算法规则给出了单值选择函数的和型Fuzzy实现的有效算法;最后还给出了一个具体例子。
In this paper, some problems about the sum-fuzzy implementation of singlevalued choice functions are investigated based on [1]. This kind of problems is first changed into the model of perceptrons in neural theory, Under the assumption of a finite set of social states. a sufficient and necessary condition for single-valued choice functions to be sum-fuzzy implementable is obtained: a geometric characterization is also given: an efficient algorithm for the sumfuzzy implementation of single-valued choice functions is obtained based on the rules of learning algorithm of perceptrons. Finally, a concrete example is given.
出处
《管理工程学报》
CSSCI
1993年第3期171-174,共4页
Journal of Industrial Engineering and Engineering Management
关键词
选择函数
和型模糊实现
神经元
Choice functions
sum—fuzzy rational
Perceptrona
Learning algorithm
