Machine-learning-assisted Monte Carlo fails at sampling computationally hard problems
Machine Learning: Science and Technology · 01.03.2023 · 4 (1) · 010501
DOI: 10.1088/2632-2153/acbe91 ↗Kurz gefasst
Fachartikel in der Zeitschrift Machine Learning: Science and Technology, online erschienen am 14.03.2023. Verfasst von Simone Ciarella und 3 weiteren Autor:innen. Die Arbeit wurde bisher 32-mal zitiert (Crossref) und nennt 57 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: H2020 European Research Council.
Abstract (Originalsprache)
Several strategies have been recently proposed in order to improve Monte Carlo sampling efficiency using machine learning tools. Here, we challenge these methods by considering a class of problems that are known to be exponentially hard to sample using conventional local Monte Carlo at low enough temperatures. In particular, we study the antiferromagnetic Potts model on a random graph, which reduces to the coloring of random graphs at zero temperature. We test several machine-learning-assisted Monte Carlo approaches, and we find that they all fail. Our work thus provides good benchmarks for future proposals for smart sampling algorithms.
Förderung
- H2020 European Research Council
Zitieren
APA
Ciarella, S., Trinquier, J., Weigt, M., & Zamponi, F. (2023). Machine-learning-assisted Monte Carlo fails at sampling computationally hard problems. Machine Learning: Science and Technology, 4(1), 010501. https://doi.org/10.1088/2632-2153/acbe91
Vancouver
Ciarella S, Trinquier J, Weigt M, Zamponi F. Machine-learning-assisted Monte Carlo fails at sampling computationally hard problems. Machine Learning: Science and Technology. 2023;4(1):010501. doi:10.1088/2632-2153/acbe91
BibTeX
@article{Ciarella2023,
author = {Ciarella, Simone and Trinquier, Jeanne and Weigt, Martin and Zamponi, Francesco},
title = {Machine-learning-assisted Monte Carlo fails at sampling computationally hard problems},
journal = {Machine Learning: Science and Technology},
year = {2023},
volume = {4},
number = {1},
pages = {010501},
doi = {10.1088/2632-2153/acbe91}
}
EUVIMED