A cautionary tale for machine learning generated configurations in presence of a conserved quantity

Abstract We investigate the performance of machine learning algorithms trained exclusively with configurations obtained from importance sampling Monte Carlo simulations of the two-dimensional Ising model with conserved magnetization. For supervised machine learning, we use convolutional neural netwo...

Descripción completa

Guardado en:

Detalles Bibliográficos
Autores principales:	Ahmadreza Azizi, Michel Pleimling
Formato:	article
Lenguaje:	EN
Publicado:	Nature Portfolio 2021
Materias:	Medicine R Science Q
Acceso en línea:	https://doaj.org/article/df1580ca436345bc98dcc70bfc4d5ba2
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

id	oai:doaj.org-article:df1580ca436345bc98dcc70bfc4d5ba2
record_format	dspace
spelling	oai:doaj.org-article:df1580ca436345bc98dcc70bfc4d5ba22021-12-02T13:17:48ZA cautionary tale for machine learning generated configurations in presence of a conserved quantity10.1038/s41598-021-85683-82045-2322https://doaj.org/article/df1580ca436345bc98dcc70bfc4d5ba22021-03-01T00:00:00Zhttps://doi.org/10.1038/s41598-021-85683-8https://doaj.org/toc/2045-2322Abstract We investigate the performance of machine learning algorithms trained exclusively with configurations obtained from importance sampling Monte Carlo simulations of the two-dimensional Ising model with conserved magnetization. For supervised machine learning, we use convolutional neural networks and find that the corresponding output not only allows to locate the phase transition point with high precision, it also displays a finite-size scaling characterized by an Ising critical exponent. For unsupervised learning, restricted Boltzmann machines (RBM) are trained to generate new configurations that are then used to compute various quantities. We find that RBM generates configurations with magnetizations and energies forbidden in the original physical system. The RBM generated configurations result in energy density probability distributions with incorrect weights as well as in wrong spatial correlations. We show that shortcomings are also encountered when training RBM with configurations obtained from the non-conserved Ising model.Ahmadreza AziziMichel PleimlingNature PortfolioarticleMedicineRScienceQENScientific Reports, Vol 11, Iss 1, Pp 1-10 (2021)
institution	DOAJ
collection	DOAJ
language	EN
topic	Medicine R Science Q
spellingShingle	Medicine R Science Q Ahmadreza Azizi Michel Pleimling A cautionary tale for machine learning generated configurations in presence of a conserved quantity
description	Abstract We investigate the performance of machine learning algorithms trained exclusively with configurations obtained from importance sampling Monte Carlo simulations of the two-dimensional Ising model with conserved magnetization. For supervised machine learning, we use convolutional neural networks and find that the corresponding output not only allows to locate the phase transition point with high precision, it also displays a finite-size scaling characterized by an Ising critical exponent. For unsupervised learning, restricted Boltzmann machines (RBM) are trained to generate new configurations that are then used to compute various quantities. We find that RBM generates configurations with magnetizations and energies forbidden in the original physical system. The RBM generated configurations result in energy density probability distributions with incorrect weights as well as in wrong spatial correlations. We show that shortcomings are also encountered when training RBM with configurations obtained from the non-conserved Ising model.
format	article
author	Ahmadreza Azizi Michel Pleimling
author_facet	Ahmadreza Azizi Michel Pleimling
author_sort	Ahmadreza Azizi
title	A cautionary tale for machine learning generated configurations in presence of a conserved quantity
title_short	A cautionary tale for machine learning generated configurations in presence of a conserved quantity
title_full	A cautionary tale for machine learning generated configurations in presence of a conserved quantity
title_fullStr	A cautionary tale for machine learning generated configurations in presence of a conserved quantity
title_full_unstemmed	A cautionary tale for machine learning generated configurations in presence of a conserved quantity
title_sort	cautionary tale for machine learning generated configurations in presence of a conserved quantity
publisher	Nature Portfolio
publishDate	2021
url	https://doaj.org/article/df1580ca436345bc98dcc70bfc4d5ba2
work_keys_str_mv	AT ahmadrezaazizi acautionarytaleformachinelearninggeneratedconfigurationsinpresenceofaconservedquantity AT michelpleimling acautionarytaleformachinelearninggeneratedconfigurationsinpresenceofaconservedquantity AT ahmadrezaazizi cautionarytaleformachinelearninggeneratedconfigurationsinpresenceofaconservedquantity AT michelpleimling cautionarytaleformachinelearninggeneratedconfigurationsinpresenceofaconservedquantity
_version_	1718393360168779776

A cautionary tale for machine learning generated configurations in presence of a conserved quantity

Ejemplares similares