Robust wave function optimization procedures in quantum Monte Carlo methods

Abstract

The energy variance optimization algorithm over a fixed ensemble of configurations in variational Monte Carlo is formally identical to a problem of fitting data: we reexamine it from a statistical maximum-likelihood point of view. We detect the origin of the problem of convergence that is often encountered in practice and propose an alternative procedure for optimization of trial wave functions in quantum Monte Carlo. We successfully test this proposal by optimizing a trial wave function for the Helium trimer.

Submitted for publication

Cited by in corpus (14)

• Correlated geminal wave function for molecules: an efficient resonating valence bond approach
• Variational and Diffusion Quantum Monte Carlo Calculations with the CASINO Code
• $\sigma$-SCF: A Direct Energy-targeting Method To Mean-field Excited States
• Accurate, efficient and simple forces with Quantum Monte Carlo methods
• Riemannian geometry and automatic differentiation for optimization problems of quantum physics and quantum technologies
• Heavy-tailed random error in quantum Monte Carlo
• Predicting atomic dopant solvation in helium clusters: the MgHe${}_n$ case
• QGOpt: Riemannian optimization for quantum technologies
• Variational autoencoder reconstruction of complex many-body physics
• Optimization of inhomogeneous electron correlation factors in periodic solids
• Quantum Monte Carlo study of the phase diagram of the two-dimensional uniform electron liquid
• Delayed Rejection Variational Monte Carlo
• Decorrelation of samples in Quantum Monte Carlo calculations and scaling of autocorrelation time in Li and H${}_2$O clusters
• The two-positron gluic bond as a manifestation of "super" van der Waals interactions