Carregant...
Carregant...

Vés al contingut (premeu Retorn)

Learning by back-propagation computing in a systolic way

Autor
Millan, J.; Bofill, P.
Tipus d'activitat
Article en revista
Revista
Lecture notes in computer science
Data de publicació
1989-10
Volum
366
Pàgina inicial
235
Pàgina final
252
DOI
https://doi.org/10.1007/3-540-51285-3_44 Obrir en finestra nova
URL
https://link.springer.com/chapter/10.1007/3-540-51285-3_44 Obrir en finestra nova
Resum
In this paper we present a systolic algorithm for back-propagation, a supervised, iteratived, gradient-descent, connectionist learning rule. The algorithm works on feedforward networks where connections can skip layers and fully exploits spatial and training parallelism, which are inherent to back-propagation. Spatial parallelism arises during the propagation of activity—forward—and error—backward—for a particular input-output. On the other hand, when this computation is carried out simu...
Paraules clau
Back propagation, Spatial parallelism, Systolic algorithm
Grup de recerca
CAP - Grup de Computació d'Altes Prestacions

Participants