论文部分内容阅读
本文首先讨论了组合式前馈神经网络原理和一般设计方法， 然后用两个仿真实验展示了组合式前馈神经网络在处理函数逼近和模式分类问题上的有效性。
This paper first discusses the principle and general design method of the combined feedforward neural network, and then demonstrates the effectiveness of the combined feedforward neural network on the approximation of processing functions and the classification of modes by two simulation experiments.