【作者】 吴青娥； 舒兰；

【Author】 WU Qing-e,SHU Lan (Dept. of Applied Mathematics University of Electronic Science and technology of China, Chengdu 610054,China)

【机构】 四川成都电子科技大学应用数学学院；

【摘要】 到目前为止,我们所研究的模糊或非模糊的自动机都是有限状态自动机。然而,关于无限状态自动机的定义及它的稳定性和收敛性都没有被讨论过。本文中,我们使用离散的反馈神经网络及网络输出空间划分方法,同时,在梯度更新算法中使用伪梯度方法,给出了模糊无限状态自动机收敛到模糊有限状态自动机的证明。更多还原

【Abstract】 The automata of fuzzy or defuzzization what we study are all finite-state automata so far. However, the definition of infinite-state automaton is not introduced and the infinite-state automaton is not discussed for its stability and convergence. In this paper, fuzzy infinite-state automata that converge to fuzzy finite-state automata will be discussed by using discrete recurrent neural network and using the partition of output space of recurrent neural network, at the same time, the pseudo-gradient is used in our gradient update calculation.更多还原