No full text
Poster (Scientific congresses and symposiums)
A zealous parallel gradient descent algorithm
Louppe, Gilles; Geurts, Pierre
2010NIPS 2010 Workshop on Learning on Cores, Clusters and Clouds
 

Files


Full Text
No document available.
Full Text Parts
glouppe10-lccc-poster.pdf
Author preprint (919.42 kB)
Poster
Download
glouppe-pgeurts-lccc2010.pdf
Author preprint (104.8 kB)
Extended abstract
Download
Annexes
rbm-zealous.zip
Publisher postprint (36.41 kB)
Source code
Download

All documents in ORBi are protected by a user license.

Send to



Details



Keywords :
machine learning; optimization; gradient descent
Abstract :
[en] Parallel and distributed algorithms have become a necessity in modern machine learning tasks. In this work, we focus on parallel asynchronous gradient descent and propose a zealous variant that minimizes the idle time of processors to achieve a substantial speedup. We then experimentally study this algorithm in the context of training a restricted Boltzmann machine on a large collaborative filtering task.
Disciplines :
Computer science
Author, co-author :
Louppe, Gilles  ;  Université de Liège - ULiège > Dép. d'électric., électron. et informat. (Inst.Montefiore) > Systèmes et modélisation
Geurts, Pierre ;  Université de Liège - ULiège > Dép. d'électric., électron. et informat. (Inst.Montefiore) > Systèmes et modélisation
Language :
English
Title :
A zealous parallel gradient descent algorithm
Publication date :
11 December 2010
Event name :
NIPS 2010 Workshop on Learning on Cores, Clusters and Clouds
Event place :
Whistler, Canada
Event date :
11 décembre 2010
Audience :
International
Funders :
F.R.S.-FNRS - Fonds de la Recherche Scientifique [BE]
Available on ORBi :
since 04 January 2011

Statistics


Number of views
538 (63 by ULiège)
Number of downloads
486 (41 by ULiège)

Bibliography


Similar publications



Contact ORBi