摘要
模糊神经推理机是模糊逻辑和神经计算相结合的产物 ,特征的模糊化对神经网络的学习提出了更高的要求 ,传统的 BP算法由于自身的局限影响了推理机的性能 .本文将进化规划即 EP(Evolutionary Programming)引入这一领域 ,基于 EP的学习算法使得神经网络的结构调整更加灵活 。
Fuzzy neural reasoning machine is the result of combination of fuzzy logics and neural computing. Fuzzy character requires more about the learning of neural networks. BP algorithm affects the performance of the reasoning machine because of its own localization. This paper introduces EP(Evolutionary Programming)to that field. EP makes it more flexible adjust the neural network, so it can provides better reasoning machine.
出处
《小型微型计算机系统》
CSCD
北大核心
2003年第3期467-469,共3页
Journal of Chinese Computer Systems
基金
国防科工委基础研究项目 (980 6992 3 )资助
