摘要
本文对文献[1]《一种改进的模糊参数神经网络学习算法》中提出的F-BP算法进行分析研究,提出了一种效率更好的快速F-BP学习算法.在此算法中对模糊数的计算方法进行改进,减少了模糊神经网络训练及推理中的运算量,从而提高学习效率.
This paper improves the F-BP algorithm which was proposed in [1] 《the 1 earning algorithm of the neural network with bell--shaped fuzzy parameter》 We propose a more efficient and fast F-BP learning algorithm. In this algorithm, we improve the calculation methord of the fuzzy number, and thereby decrease the capacity of the calculation in fuzzy-neural-network training and inference , which improves the learning efficiency.
出处
《新疆大学学报（自然科学版）》
CAS
2007年第3期349-352,共4页
Journal of Xinjiang University(Natural Science Edition)
关键词
模糊神经网络
模糊推理
模糊数
fuzzy-neural-network
fuzzy inference
fuzzy number
