Correspondence between neuroevolution and gradient descent

Gradient-based and non-gradient-based methods for training neural networks are usually considered to be fundamentally different. The authors derive, and illustrate numerically, an analytic equivalence between the dynamics of neural network training under conditioned stochastic mutations, and under g...

Descripción completa

Guardado en:

Detalles Bibliográficos
Autores principales:	Stephen Whitelam, Viktor Selin, Sang-Won Park, Isaac Tamblyn
Formato:	article
Lenguaje:	EN
Publicado:	Nature Portfolio 2021
Materias:	Science Q
Acceso en línea:	https://doaj.org/article/2228d1b435c34f58901cee411ded17c8
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

id	oai:doaj.org-article:2228d1b435c34f58901cee411ded17c8
record_format	dspace
spelling	oai:doaj.org-article:2228d1b435c34f58901cee411ded17c82021-11-08T11:07:35ZCorrespondence between neuroevolution and gradient descent10.1038/s41467-021-26568-22041-1723https://doaj.org/article/2228d1b435c34f58901cee411ded17c82021-11-01T00:00:00Zhttps://doi.org/10.1038/s41467-021-26568-2https://doaj.org/toc/2041-1723Gradient-based and non-gradient-based methods for training neural networks are usually considered to be fundamentally different. The authors derive, and illustrate numerically, an analytic equivalence between the dynamics of neural network training under conditioned stochastic mutations, and under gradient descent.Stephen WhitelamViktor SelinSang-Won ParkIsaac TamblynNature PortfolioarticleScienceQENNature Communications, Vol 12, Iss 1, Pp 1-10 (2021)
institution	DOAJ
collection	DOAJ
language	EN
topic	Science Q
spellingShingle	Science Q Stephen Whitelam Viktor Selin Sang-Won Park Isaac Tamblyn Correspondence between neuroevolution and gradient descent
description	Gradient-based and non-gradient-based methods for training neural networks are usually considered to be fundamentally different. The authors derive, and illustrate numerically, an analytic equivalence between the dynamics of neural network training under conditioned stochastic mutations, and under gradient descent.
format	article
author	Stephen Whitelam Viktor Selin Sang-Won Park Isaac Tamblyn
author_facet	Stephen Whitelam Viktor Selin Sang-Won Park Isaac Tamblyn
author_sort	Stephen Whitelam
title	Correspondence between neuroevolution and gradient descent
title_short	Correspondence between neuroevolution and gradient descent
title_full	Correspondence between neuroevolution and gradient descent
title_fullStr	Correspondence between neuroevolution and gradient descent
title_full_unstemmed	Correspondence between neuroevolution and gradient descent
title_sort	correspondence between neuroevolution and gradient descent
publisher	Nature Portfolio
publishDate	2021
url	https://doaj.org/article/2228d1b435c34f58901cee411ded17c8
work_keys_str_mv	AT stephenwhitelam correspondencebetweenneuroevolutionandgradientdescent AT viktorselin correspondencebetweenneuroevolutionandgradientdescent AT sangwonpark correspondencebetweenneuroevolutionandgradientdescent AT isaactamblyn correspondencebetweenneuroevolutionandgradientdescent
_version_	1718442312838676480

Correspondence between neuroevolution and gradient descent

Ejemplares similares