摘要
本文介绍一种新的前馈神经网络的随机学习方法，着重讨论该算法的实现，并讨论了将它和ＢＰ算法相结合，从而得到一种非常实用的神经网络学习算法。
In this paper, we introduce a statistical training algorithm for feedforward artificial neural networks, especially in the realization of the algorithm. We also discussed the combination of this method and BP algorithm, then we can get a practical and original learning algorithm for the Artificial Neural Networks.
出处
《微机发展》
1998年第3期4-6,共3页
Microcomputer Development
关键词
前馈神经网络
学习算法
复合形法
BP算法
Feedforward Artificial Neural Networks
Training Algorithm
Complex Method
Back Propagation Algorithm
Statistical Methods
Supervised Training
Local Minima
