We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

physics.comp-ph

Change to browse by:

References & Citations

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Physics > Computational Physics

Title: Deep learning density functionals for gradient descent optimization

Abstract: Machine-learned regression models represent a promising tool to implement accurate and computationally affordable energy-density functionals to solve quantum many-body problems via density functional theory. However, while they can easily be trained to accurately map ground-state density profiles to the corresponding energies, their functional derivatives often turn out to be too noisy, leading to instabilities in self-consistent iterations and in gradient-based searches of the ground-state density profile. We investigate how these instabilities occur when standard deep neural networks are adopted as regression models, and we show how to avoid it using an ad-hoc convolutional architecture featuring an inter-channel averaging layer. The testbed we consider is a realistic model for noninteracting atoms in optical speckle disorder. With the inter-channel average, accurate and systematically improvable ground-state energies and density profiles are obtained via gradient-descent optimization, without instabilities nor violations of the variational principle.
Comments: 9 pages, 10 figures
Subjects: Computational Physics (physics.comp-ph); Disordered Systems and Neural Networks (cond-mat.dis-nn); Materials Science (cond-mat.mtrl-sci); Other Condensed Matter (cond-mat.other); Quantum Gases (cond-mat.quant-gas)
Journal reference: Phys. Rev. E 106, 045309 (2022)
DOI: 10.1103/PhysRevE.106.045309
Cite as: arXiv:2205.08367 [physics.comp-ph]
  (or arXiv:2205.08367v2 [physics.comp-ph] for this version)

Submission history

From: Emanuele Costa [view email]
[v1] Tue, 17 May 2022 13:57:08 GMT (486kb,D)
[v2] Mon, 7 Nov 2022 12:14:42 GMT (120kb,D)

Link back to: arXiv, form interface, contact.