论文部分内容阅读
本文从控制系统的角度讨论了线性和型人工神经网络的拓扑结构和统一描述问题,对一些典型网络模型的分析表明:采用控制系统理论中关于系统描述与结构分析的传递矩阵和框图方法来分析神经网络,不仅可使网络的结构清晰化,而且还可为网络的综合提供一条系统的思路.
This paper discusses the topological structure and uniform description of linear and artificial neural networks from the point of view of control system. The analysis of some typical network models shows that the transfer matrix and block diagram method of system description and structure analysis are used to analyze Neural network can not only make the structure of the network clear, but also provide a systematic idea for the network synthesis.