Parallel recursive prediction error algorithm for training layered neural networks

A new recursive prediction error algorithm is derived for the training of feedforward layered neural networks. The algorithm enables the weights in each neuron of the network to be updated in an efficient parallel manner and has better convergence properties than the classical back propagation algor...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:International journal of control Ročník 51; číslo 6; s. 1215 - 1228
Hlavní autoři: CHEN, S., COWAN, C. F. N., BILLINGS, S. A., GRANT, P. M.
Médium: Journal Article
Jazyk:angličtina
Vydáno: London Taylor & Francis Group 1990
Taylor & Francis
Témata:
Applied sciences
Computer science; control theory; systems
Control theory. Systems
Exact sciences and technology
Modelling and identification
Backpropagation
Parallel algorithm
Steepest descent method
Gauss Newton method
Prediction
Data transmission
Error
Neural network
Equalization
Non linear system
Recursive algorithm
Modeling
Digital transmission
SISO system
System identification
ISSN:0020-7179, 1366-5820
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Popis
Shrnutí:A new recursive prediction error algorithm is derived for the training of feedforward layered neural networks. The algorithm enables the weights in each neuron of the network to be updated in an efficient parallel manner and has better convergence properties than the classical back propagation algorithm. The relationship between this new parallel algorithm and other existing learning algorithms is discussed. Examples taken from the fields of communication channel equalization and nonlinear systems modelling are used to demonstrate the superior performance of the new algorithm compared with the back propagation routine.
Bibliografie:ObjectType-Article-2
SourceType-Scholarly Journals-1
ObjectType-Feature-1
content type line 23
ISSN:0020-7179
1366-5820
DOI:10.1080/00207179008934127