Learning Kinetic Monte Carlo stochastic dynamics with Deep Generative Adversarial Networks

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Lanzoni, Daniele, Pierre-Louis, Olivier, Bergamaschini, Roberto, Montalenti, Francesco
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866911081736175616
author Lanzoni, Daniele
Pierre-Louis, Olivier
Bergamaschini, Roberto
Montalenti, Francesco
author_facet Lanzoni, Daniele
Pierre-Louis, Olivier
Bergamaschini, Roberto
Montalenti, Francesco
contents We show that Generative Adversarial Networks (GANs) may be fruitfully exploited to learn stochastic dynamics, surrogating traditional models while capturing thermal fluctuations. Specifically, we showcase the application to a two-dimensional, many-particle system, focusing on surface-step fluctuations and on the related time-dependent roughness. After the construction of a dataset based on Kinetic Monte Carlo simulations, a conditional GAN is trained to propagate stochastically the state of the system in time, allowing the generation of new sequences with a reduced computational cost. Modifications with respect to standard GANs, which facilitate convergence and increase accuracy, are discussed. The trained network is demonstrated to quantitatively reproduce equilibrium and kinetic properties, including scaling laws, with deviations of a few percent from the exact value. Extrapolation limits and future perspectives are critically discussed.
format Preprint
id arxiv_https___arxiv_org_abs_2507_21763
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Learning Kinetic Monte Carlo stochastic dynamics with Deep Generative Adversarial Networks
Lanzoni, Daniele
Pierre-Louis, Olivier
Bergamaschini, Roberto
Montalenti, Francesco
Statistical Mechanics
Materials Science
Artificial Intelligence
Machine Learning
Computational Physics
We show that Generative Adversarial Networks (GANs) may be fruitfully exploited to learn stochastic dynamics, surrogating traditional models while capturing thermal fluctuations. Specifically, we showcase the application to a two-dimensional, many-particle system, focusing on surface-step fluctuations and on the related time-dependent roughness. After the construction of a dataset based on Kinetic Monte Carlo simulations, a conditional GAN is trained to propagate stochastically the state of the system in time, allowing the generation of new sequences with a reduced computational cost. Modifications with respect to standard GANs, which facilitate convergence and increase accuracy, are discussed. The trained network is demonstrated to quantitatively reproduce equilibrium and kinetic properties, including scaling laws, with deviations of a few percent from the exact value. Extrapolation limits and future perspectives are critically discussed.
title Learning Kinetic Monte Carlo stochastic dynamics with Deep Generative Adversarial Networks
topic Statistical Mechanics
Materials Science
Artificial Intelligence
Machine Learning
Computational Physics
url https://arxiv.org/abs/2507.21763