Improving dynamical properties of metropolized discretizations of overdamped Langevin dynamics
For researchers in computational statistical physics and molecular dynamics, this provides incremental improvements to dynamical consistency of Metropolis-corrected Langevin integrators.
This work modifies the Metropolis-Hastings correction for discretizations of overdamped Langevin dynamics on compact spaces to improve either the strong order of the numerical method or reduce bias in transport coefficient estimation, using modified schemes and a Barker acceptance rule.
The discretization of overdamped Langevin dynamics, through schemes such as the Euler-Maruyama method, can be corrected by some acceptance/rejection rule, based on a Metropolis-Hastings criterion for instance. In this case, the invariant measure sampled by the Markov chain is exactly the Boltzmann-Gibbs measure. However, rejections perturb the dynamical consistency of the resulting numerical method with the reference dynamics. We present in this work some modifications of the standard correction of discretizations of overdamped Langevin dynamics on compact spaces by a Metropolis-Hastings procedure, which allow us to either improve the strong order of the numerical method, or to decrease the bias in the estimation of transport coefficients characterizing the effective dynamical behavior of the dynamics. For the latter approach, we rely on modified numerical schemes together with a Barker rule for the acceptance/rejection criterion.