摘要
对日前应用最广的人工神经网络——BP网络的学习算法及各种改进学习算法进行了分类比较和综合分析,并结合作者的研究结果对进一步研究BP网络及算法进行了探讨.
In this paper, a comprehensive analysis by classification is carried out on current representative improved learning algorithms for BP net, one of the most widely used Artificial Neural Networks. With authors' research results taken into consideration , some suggestions on the further improvement in BP algorithms are proposed.
出处
《应用基础与工程科学学报》
EI
CSCD
1995年第4期105-110,共6页
Journal of Basic Science and Engineering
关键词
人工神经网络
误差反向传播
隐节点
机器学习
artificial neural networks, error back propagation, hidden node, machine learning
