论文部分内容阅读
本文旨在探讨符号序列在人工神经网络中的记忆、表达和回忆的机理。通过利用符号的音频表示、环境反馈网络、符号动力系统和激发序列等理论概念建造具体模型,着重分析了简单动力系统中的激发序列在神经网络系统中的记忆、表达和回忆的过程。本文有助于促进我们去研究层次结构更为复杂的符号序列在神经网络系统中的统一表达形式,也有助于我们去探讨人类记忆的本质。
The purpose of this paper is to explore the mechanism of symbolic sequences in memory, expression and recall in artificial neural networks. The specific model is constructed through the use of symbolic audio representation, environmental feedback network, symbolic dynamical system and excitation sequence, and emphatically analyzes the process of memory, expression and recall of excitation sequences in neural networks. This article helps us to study the unified expression of symbolic sequences with more complex hierarchies in neural network systems and also helps to explore the essence of human memory.