HMC: avoiding rejections by not using leapfrog and some results on the acceptance rate
From MaRDI portal
Publication:6330601
DOI10.1016/J.JCP.2021.110333arXiv1912.03253MaRDI QIDQ6330601FDOQ6330601
Authors: M. Calvo, Daniel Sanz-Alonso, Jesús María Sanz-Serna 
Publication date: 6 December 2019
Abstract: The leapfrog integrator is routinely used within the Hamiltonian Monte Carlo method and its variants. We give strong numerical evidence that alternative, easy to implement algorithms yield fewer rejections with a given computational effort. When the dimensionality of the target distribution is high, the number of accepted proposals may be multiplied by a factor of three or more. This increase in the number of accepted proposals is not achieved by impairing any positive features of the sampling. We also establish new non-asymptotic and asymptotic results on the monotonic relationship between the expected acceptance rate and the expected energy error. These results further validate the derivation of one of the integrators we consider and are of independent interest.
Numerical methods for ordinary differential equations (65Lxx) Probabilistic methods, stochastic differential equations (65Cxx) Markov processes (60Jxx)
This page was built for publication: HMC: avoiding rejections by not using leapfrog and some results on the acceptance rate
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6330601)
