The backpropagation unfold recurrent rule
暂无分享,去创建一个
A novel learning algorithm for recurrent backpropagation (BP) networks is introduced. Existing learning algorithms of the recurrent models are mostly based on the static generalized delta-rule of the BP algorithm which does not give satisfactory results for a recurrent network. The authors derived the BP unfold recurrent rule (BURR) by unfolding the structure of the recurrent network. Using this learning rule, the connecting weights are updated by accumulating the effects of the weight change in each time interval. Experimental results are presented to demonstrate the memorizing ability of a recurrent network using this learning algorithm. Sequences of input characters were successfully recognized.<<ETX>>
[1] Geoffrey E. Hinton,et al. Learning internal representations by error propagation , 1986 .
[2] Frank Fallside,et al. An adaptive training algorithm for back propagation networks , 1987 .
[3] Stefano Noffi,et al. Recall of Sequences of Items by a Neural Network , 1991 .
[4] J J Hopfield,et al. Neural computation by concentrating information in time. , 1987, Proceedings of the National Academy of Sciences of the United States of America.