Frontiers in Physics (Jun 2023)

Efficient solutions of fermionic systems using artificial neural networks

  • Even M. Nordhagen,
  • Jane M. Kim,
  • Bryce Fore,
  • Alessandro Lovato,
  • Morten Hjorth-Jensen,
  • Morten Hjorth-Jensen

DOI
https://doi.org/10.3389/fphy.2023.1061580
Journal volume & issue
Vol. 11

Abstract

Read online

In this study, we explore the similarities and differences between variational Monte Carlo techniques that employ conventional and artificial neural network representations of the ground-state wave function for fermionic systems. Our primary focus is on shallow neural network architectures, specifically the restricted Boltzmann machine, and we examine unsupervised learning algorithms that are appropriate for modeling complex many-body correlations. We assess the advantages and drawbacks of conventional and neural network wave functions by applying them to a range of circular quantum dot systems. Our findings, which include results for systems containing up to 90 electrons, emphasize the efficient implementation of these methods on both homogeneous and heterogeneous high-performance computing facilities.

Keywords