A machine-learning based approach to the evaluation of the critical scaling behavior of anisotropic spin systems

A machine-learning based approach to the evaluation of the critical scaling behavior of anisotropic spin systems Full article

Journal

Journal of Physics: Condensed Matter
ISSN: 1361-648X

Output data

Year: 2026, Volume: 38, Number: 39, Article number : 395403, Pages count : 15 DOI: 10.1088/1361-648X/aea9de

Tags

critical phenomena, phase transitions, anisotropic heisenberg model, convolutional neural networks, finite-size scaling, monte carlo simulations

Authors

Chubarova Alina A ¹ , Mamonov Ivan A ² , Mamonova Marina V ² , Bogachev Mikhail I ³ , Prudnikov Pavel V ¹

Affiliations

1	Center of New Chemical Technologies BIC, Boreskov Institute of Catalysis
2	Theoretical Physics Department, Dostoevsky Omsk State University
3	Faculty of Radio Engineering & Telecommunication, St. Petersburg State Electrotechnical University ‘LETI’

Funding (1)

Ministry of Science and Higher Education of the Russian Federation

FWUR-2024-0039

Computational models adequately representing phase transitions and evaluating the critical system parameters are essential for the understanding of the properties of a wide range of materials. Here we propose a machine learning (ML)-based approach to the identification of the critical point in anisotropic spin systems. Our approach implies training of a convolutional neural network (CNN) model from the correlation matrices obtained by Monte Carlo simulations. Next, the pretrained model is employed as a fast estimator of the critical temperature, which can be extracted in several complementary ways from the CNN model inference, this way improving the robustness of the analysis. The ML-based estimates obtained in this study are in very good agreement with the reference Monte Carlo simulation results, while computational costs are about 10x lower compared to the classical thermodynamic approach.

Chubarova A.A. , Mamonov I.A. , Mamonova M.V. , Bogachev M.I. , Prudnikov P.V.
A machine-learning based approach to the evaluation of the critical scaling behavior of anisotropic spin systems
Journal of Physics: Condensed Matter. 2026. V.38. N39. 395403 :1-15. DOI: 10.1088/1361-648X/aea9de

Submitted:	Jul 17, 2026
Accepted:	Sep 18, 2026
Published online:	Sep 29, 2026

No identifiers