A NEW MODEL OF FEEDFORWARD NEURAL NETWORKS

	A NEW MODEL OF FEEDFORWARD NEURAL NETWORKS
	WANG, DX; TAI, JW
刊名	PHYSICS LETTERS A
	1992-01-27
卷号	162 期号:1 页码:41-44
文献子类	Article
英文摘要	A new model of multilayer neural networks based on the back-propagation algorithm is presented. The network is a modified version of the STBP model, and employs an output layer with hard threshold function neurons; the model is very efficient; the learning speed can be improved. The model has an efficient ability to approximate any practical function, and is suitable for optical implementation.
WOS研究方向	Physics
语种	英语
WOS记录号	WOS:A1992HE83200012
内容类型	期刊论文
源URL	[http://ir.ia.ac.cn/handle/173211/21885]
专题	自动化研究所_09年以前成果
推荐引用方式 GB/T 7714	WANG, DX,TAI, JW. A NEW MODEL OF FEEDFORWARD NEURAL NETWORKS[J]. PHYSICS LETTERS A,1992,162(1):41-44.
APA	WANG, DX,&TAI, JW.(1992).A NEW MODEL OF FEEDFORWARD NEURAL NETWORKS.PHYSICS LETTERS A,162(1),41-44.
MLA	WANG, DX,et al."A NEW MODEL OF FEEDFORWARD NEURAL NETWORKS".PHYSICS LETTERS A 162.1(1992):41-44.