EUVIMEDEUVIMED entdecken
Crossref

Second-order optimization strategies for neural network quantum states

M Drissi iD, J W T Keeble iD, J Rozalén Sarmiento iD, A Rios iD

Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences · 23.07.2024 · 382 (2275)

DOI: 10.1098/rsta.2024.0057 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences, online erschienen am 24.06.2024. Verfasst von M Drissi und 3 weiteren Autor:innen. Beteiligte Einrichtung: TRIUMF , Vancouver, British Columbia V6T 2A3, Canada. Die Arbeit wurde bisher 7-mal zitiert (Crossref) und nennt 80 Literaturangaben. Gefördert von: Science and Technology Facilities Council, Ministerio de Ciencia e Innovación, National Research Council Canada.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

The Variational Monte Carlo (VMC) method has recently seen important advances through the use of neural network quantum states. While more and more sophisticated ansatze have been designed to tackle a wide variety of quantum many-body problems, modest progress has been made on the associated optimization algorithms. In this work, we revisit the Kronecker-Factored Approximate Curvature (KFAC), an optimizer that has been used extensively in a variety of simulations. We suggest improvements in the scaling and the direction of this optimizer and find that they substantially increase its performance at a negligible additional cost. We also reformulate the VMC approach in a game theory framework, to propose a novel optimizer based on decision geometry. We find that on a practical test case for continuous systems, this new optimizer consistently outperforms any of the KFAC improvements in terms of stability, accuracy and speed of convergence. Beyond VMC, the versatility of this approach suggests that decision geometry could provide a solid foundation for accelerating a broad class of machine learning algorithms. This article is part of the theme issue 'The liminal position of Nuclear Physics: from hadrons to neutron stars'.

Schlagwörter

Stichwörter der Autor:innen

Variational Monte Carlodecision geometrygame theoryneural network quantum statesoptimization

Autor:innen und Institutionen

  • TRIUMF , Vancouver, British Columbia V6T 2A3, Canada.
  • Department of Physics, University of Surrey , Guildford, GU2 7XH, UK.
  • Departament de Física Quàntica i Astrofísica, Universitat de Barcelona (UB) , c. Martí i Franquès 1, Barcelona E08028, Spain.
  • Institut de Ciències del Cosmos (ICCUB), Universitat de Barcelona (UB) , Barcelona, Spain.

ORCID-Profile: M Drissi, J W T Keeble, J Rozalén Sarmiento, A Rios

Förderung

  • Science and Technology Facilities Council
  • Ministerio de Ciencia e Innovación
  • National Research Council Canada

Zitieren

APA

Drissi, M., Keeble, J. W. T., Rozalén Sarmiento, J., & Rios, A. (2024). Second-order optimization strategies for neural network quantum states. Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences, 382(2275). https://doi.org/10.1098/rsta.2024.0057

Vancouver

Drissi M, Keeble JWT, Rozalén Sarmiento J, Rios A. Second-order optimization strategies for neural network quantum states. Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences. 2024;382(2275). doi:10.1098/rsta.2024.0057

BibTeX
@article{Drissi2024,
  author = {Drissi, M and Keeble, J W T and Rozalén Sarmiento, J and Rios, A},
  title = {Second-order optimization strategies for neural network quantum states},
  journal = {Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences},
  year = {2024},
  volume = {382},
  number = {2275},
  doi = {10.1098/rsta.2024.0057}
}
Kostenlos starten →