Machine-learning-assisted Monte Carlo fails at sampling computationally hard problems
In: Machine Learning: Science and Technology, Jg. 4 (2023), Heft 1, S. 010501-10501
Online
academicJournal
Zugriff:
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.
Titel: |
Machine-learning-assisted Monte Carlo fails at sampling computationally hard problems
|
---|---|
Autor/in / Beteiligte Person: | Ciarella, Simone ; Trinquier, Jeanne ; Weigt, Martin ; Zamponi, Francesco |
Link: | |
Zeitschrift: | Machine Learning: Science and Technology, Jg. 4 (2023), Heft 1, S. 010501-10501 |
Veröffentlichung: | IOP Publishing, 2023 |
Medientyp: | academicJournal |
ISSN: | 2632-2153 (print) |
DOI: | 10.1088/2632-2153/acbe91 |
Schlagwort: |
|
Sonstiges: |
|