Faster Hamiltonian Monte Carlo by Learning Leapfrog Scale
Wu, Changye; Stoehr, Julien; Robert, Christian P. (2019), Faster Hamiltonian Monte Carlo by Learning Leapfrog Scale. https://basepub.dauphine.fr/handle/123456789/18740
TypeDocument de travail / Working paper
External document linkhttps://hal.archives-ouvertes.fr/hal-01968795
Series titleCahier de recherche CEREMADE, Université Paris-Dauphine
MetadataShow full item record
Abstract (EN)Hamiltonian Monte Carlo samplers have become standard algorithms for MCMC implementations, as opposed to more basic versions, but they still require some amount of tuning and calibration. Exploiting the U-turn criterion of the NUTS algorithm (Hoffman and Gelman, 2014), we propose a version of HMC that relies on the distribution of the integration time of the associated leapfrog integrator. Using in addition the primal-dual averaging method for tuning the step size of the integrator, we achieve an essentially calibration free version of HMC. When compared with the original NUTS on several benchmarks, this algorithm exhibits a significantly improved efficiency.
Subjects / KeywordsAcceleration methods, No-U-Turn Sampler
Showing items related by title and author.
Barthelme, Simon; Beffy, Magali; Chopin, Nicolas; Doucet, Arnaud; Jacob, Pierre E.; Johansen, Adam M.; Marin, Jean-Michel; Robert, Christian P. (2011) Document de travail / Working paper
Jacob, Pierre E.; Chopin, Nicolas; Robert, Christian P.; Rue, Havard (2009) Document de travail / Working paper