Cargando…

Fundamental limits to learning closed-form mathematical models from data

Given a finite and noisy dataset generated with a closed-form mathematical model, when is it possible to learn the true generating model from the data alone? This is the question we investigate here. We show that this model-learning problem displays a transition from a low-noise phase in which the t...

Descripción completa

Detalles Bibliográficos
Autores principales: Fajardo-Fontiveros, Oscar, Reichardt, Ignasi, De Los Ríos, Harry R., Duch, Jordi, Sales-Pardo, Marta, Guimerà, Roger
Formato: Online Artículo Texto
Lenguaje:English
Publicado: Nature Publishing Group UK 2023
Materias:
Acceso en línea:https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9950473/
https://www.ncbi.nlm.nih.gov/pubmed/36823107
http://dx.doi.org/10.1038/s41467-023-36657-z
_version_ 1784893171681984512
author Fajardo-Fontiveros, Oscar
Reichardt, Ignasi
De Los Ríos, Harry R.
Duch, Jordi
Sales-Pardo, Marta
Guimerà, Roger
author_facet Fajardo-Fontiveros, Oscar
Reichardt, Ignasi
De Los Ríos, Harry R.
Duch, Jordi
Sales-Pardo, Marta
Guimerà, Roger
author_sort Fajardo-Fontiveros, Oscar
collection PubMed
description Given a finite and noisy dataset generated with a closed-form mathematical model, when is it possible to learn the true generating model from the data alone? This is the question we investigate here. We show that this model-learning problem displays a transition from a low-noise phase in which the true model can be learned, to a phase in which the observation noise is too high for the true model to be learned by any method. Both in the low-noise phase and in the high-noise phase, probabilistic model selection leads to optimal generalization to unseen data. This is in contrast to standard machine learning approaches, including artificial neural networks, which in this particular problem are limited, in the low-noise phase, by their ability to interpolate. In the transition region between the learnable and unlearnable phases, generalization is hard for all approaches including probabilistic model selection.
format Online
Article
Text
id pubmed-9950473
institution National Center for Biotechnology Information
language English
publishDate 2023
publisher Nature Publishing Group UK
record_format MEDLINE/PubMed
spelling pubmed-99504732023-02-25 Fundamental limits to learning closed-form mathematical models from data Fajardo-Fontiveros, Oscar Reichardt, Ignasi De Los Ríos, Harry R. Duch, Jordi Sales-Pardo, Marta Guimerà, Roger Nat Commun Article Given a finite and noisy dataset generated with a closed-form mathematical model, when is it possible to learn the true generating model from the data alone? This is the question we investigate here. We show that this model-learning problem displays a transition from a low-noise phase in which the true model can be learned, to a phase in which the observation noise is too high for the true model to be learned by any method. Both in the low-noise phase and in the high-noise phase, probabilistic model selection leads to optimal generalization to unseen data. This is in contrast to standard machine learning approaches, including artificial neural networks, which in this particular problem are limited, in the low-noise phase, by their ability to interpolate. In the transition region between the learnable and unlearnable phases, generalization is hard for all approaches including probabilistic model selection. Nature Publishing Group UK 2023-02-24 /pmc/articles/PMC9950473/ /pubmed/36823107 http://dx.doi.org/10.1038/s41467-023-36657-z Text en © The Author(s) 2023 https://creativecommons.org/licenses/by/4.0/Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/ (https://creativecommons.org/licenses/by/4.0/) .
spellingShingle Article
Fajardo-Fontiveros, Oscar
Reichardt, Ignasi
De Los Ríos, Harry R.
Duch, Jordi
Sales-Pardo, Marta
Guimerà, Roger
Fundamental limits to learning closed-form mathematical models from data
title Fundamental limits to learning closed-form mathematical models from data
title_full Fundamental limits to learning closed-form mathematical models from data
title_fullStr Fundamental limits to learning closed-form mathematical models from data
title_full_unstemmed Fundamental limits to learning closed-form mathematical models from data
title_short Fundamental limits to learning closed-form mathematical models from data
title_sort fundamental limits to learning closed-form mathematical models from data
topic Article
url https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9950473/
https://www.ncbi.nlm.nih.gov/pubmed/36823107
http://dx.doi.org/10.1038/s41467-023-36657-z
work_keys_str_mv AT fajardofontiverososcar fundamentallimitstolearningclosedformmathematicalmodelsfromdata
AT reichardtignasi fundamentallimitstolearningclosedformmathematicalmodelsfromdata
AT delosriosharryr fundamentallimitstolearningclosedformmathematicalmodelsfromdata
AT duchjordi fundamentallimitstolearningclosedformmathematicalmodelsfromdata
AT salespardomarta fundamentallimitstolearningclosedformmathematicalmodelsfromdata
AT guimeraroger fundamentallimitstolearningclosedformmathematicalmodelsfromdata