Answers are grounded in this paper and cite their sources.
In recent papers it has been demonstrated that sampling a Gibbs distribution from an time-irreversible Langevin process is,from several points of view, when compared to sampling from a time-reversible one. Adding an appropriate drift to the overdamped Langevin equation results in a larger large rate function for the empirical measure of the process, a smaller variance for long time average of observables of the process, as well as a larger spectral gap. this work, we concentrate on irreversible Langevin samplers with a drift of increasing. The asymptotic variance is monotonically decreasing with respect to the growth the drift and we characterize its limiting behavior. For a Gibbs measure whose has one or more critical points, adding a large irreversible drift results in a of the process in a slow and fast component with fast motion along the sets of the potential and slow motion in the orthogonal direction. This result helps the variance reduction, which can be explained at the process level by the fast motion of the process along the level sets of the potential. Correspondingly limit of the asymptotic variance is the asymptotic variance of the limiting slow which is a diffusion process on a graph.
No takes yet. Share an insight, caveat, or question.
Rey-Bellet et al. (2015) studied this question.
Synapse has enriched 6 closely related papers on similar clinical questions. Consider them for comparative context: