摘要
讨论了一种模糊神经网络实现记忆的条件和记忆的特点 ,并给出了样本组格子点分布的概念 .通过定理 ,证明了在这种情况下 ,样本组可被模糊神经网络所记忆 ;证明了若样本组前n - 1个样本和整个样本组的模糊矩阵的秩相等 ,则权向量不能调整 ,否则会使模糊神经网络“丧失记忆” .同时 ,也说明了选择记忆法所采用的选择方法既可减少运算量 ,又可保证网络的记忆 .
The condition for fuzzy neural network to realize memory is discussed, and the concept of lattice point is given. It has been proved that if example set is of lattice point distribution the example set can be memorized by a fuzzy neural network, otherwise the fuzzy neural netowrk will lost its memory. A theorem is proposed to show that choice-memory method not only simplifies computation but also ensures memory of fuzzy neural network.
出处
《西安交通大学学报》
EI
CAS
CSCD
北大核心
2001年第2期171-174,共4页
Journal of Xi'an Jiaotong University
基金
国家"八六三"计划资助!项目 (5 1- 945- 0 11) .
关键词
模糊神经网络
格子点
记忆
Control systems
Data storage equipment
Fuzzy sets
Nonlinear systems
Theorem proving
