Matches in SemOpenAlex for { <https://semopenalex.org/work/W2542623543> ?p ?o ?g. }
Showing items 1 to 94 of
94
with 100 items per page.
- W2542623543 abstract "We introduce a new family of MCMC samplers that combine auxiliary variables, Gibbs sampling and Taylor expansions of the target density. Our approach permits the marginalisation over the auxiliary variables yielding marginal samplers, or the augmentation of the auxiliary variables, yielding auxiliary samplers. The well-known Metropolis-adjusted Langevin algorithm (MALA) and preconditioned Crank-Nicolson Langevin (pCNL) algorithm are shown to be special cases. We prove that marginal samplers are superior in terms of asymptotic variance and demonstrate cases where they are slower in computing time compared to auxiliary samplers. In the context of latent Gaussian models we propose new auxiliary and marginal samplers whose implementation requires a single tuning parameter, which can be found automatically during the transient phase. Extensive experimentation shows that the increase in efficiency (measured as effective sample size per unit of computing time) relative to (optimised implementations of) pCNL, elliptical slice sampling and MALA ranges from 10-fold in binary classification problems to 25-fold in log-Gaussian Cox processes to 100-fold in Gaussian process regression, and it is on par with Riemann manifold Hamiltonian Monte Carlo in an example where the latter has the same complexity as the aforementioned algorithms. We explain this remarkable improvement in terms of the way alternative samplers try to approximate the eigenvalues of the target. We introduce a novel MCMC sampling scheme for hyperparameter learning that builds upon the auxiliary samplers. The MATLAB code for reproducing the experiments in the article is publicly available and a Supplement to this article contains additional experiments and implementation details." @default.
- W2542623543 created "2016-11-04" @default.
- W2542623543 creator A5017255421 @default.
- W2542623543 creator A5061893818 @default.
- W2542623543 date "2016-10-30" @default.
- W2542623543 modified "2023-09-27" @default.
- W2542623543 title "Auxiliary gradient-based sampling algorithms" @default.
- W2542623543 cites W1545319692 @default.
- W2542623543 cites W1579781813 @default.
- W2542623543 cites W1954227087 @default.
- W2542623543 cites W1983452151 @default.
- W2542623543 cites W2008703230 @default.
- W2542623543 cites W2048971218 @default.
- W2542623543 cites W2055340805 @default.
- W2542623543 cites W2101517901 @default.
- W2542623543 cites W2114964853 @default.
- W2542623543 cites W2122891730 @default.
- W2542623543 cites W2124086369 @default.
- W2542623543 cites W2134117233 @default.
- W2542623543 cites W2152657433 @default.
- W2542623543 cites W2166414779 @default.
- W2542623543 cites W2167433878 @default.
- W2542623543 cites W2169410692 @default.
- W2542623543 cites W2241364617 @default.
- W2542623543 cites W2949152134 @default.
- W2542623543 hasPublicationYear "2016" @default.
- W2542623543 type Work @default.
- W2542623543 sameAs 2542623543 @default.
- W2542623543 citedByCount "1" @default.
- W2542623543 countsByYear W25426235432018 @default.
- W2542623543 crossrefType "posted-content" @default.
- W2542623543 hasAuthorship W2542623543A5017255421 @default.
- W2542623543 hasAuthorship W2542623543A5061893818 @default.
- W2542623543 hasConcept C105795698 @default.
- W2542623543 hasConcept C107673813 @default.
- W2542623543 hasConcept C111350023 @default.
- W2542623543 hasConcept C11413529 @default.
- W2542623543 hasConcept C121332964 @default.
- W2542623543 hasConcept C126255220 @default.
- W2542623543 hasConcept C140779682 @default.
- W2542623543 hasConcept C158424031 @default.
- W2542623543 hasConcept C163716315 @default.
- W2542623543 hasConcept C19499675 @default.
- W2542623543 hasConcept C28826006 @default.
- W2542623543 hasConcept C33923547 @default.
- W2542623543 hasConcept C41008148 @default.
- W2542623543 hasConcept C61326573 @default.
- W2542623543 hasConcept C62520636 @default.
- W2542623543 hasConcept C76155785 @default.
- W2542623543 hasConcept C94915269 @default.
- W2542623543 hasConceptScore W2542623543C105795698 @default.
- W2542623543 hasConceptScore W2542623543C107673813 @default.
- W2542623543 hasConceptScore W2542623543C111350023 @default.
- W2542623543 hasConceptScore W2542623543C11413529 @default.
- W2542623543 hasConceptScore W2542623543C121332964 @default.
- W2542623543 hasConceptScore W2542623543C126255220 @default.
- W2542623543 hasConceptScore W2542623543C140779682 @default.
- W2542623543 hasConceptScore W2542623543C158424031 @default.
- W2542623543 hasConceptScore W2542623543C163716315 @default.
- W2542623543 hasConceptScore W2542623543C19499675 @default.
- W2542623543 hasConceptScore W2542623543C28826006 @default.
- W2542623543 hasConceptScore W2542623543C33923547 @default.
- W2542623543 hasConceptScore W2542623543C41008148 @default.
- W2542623543 hasConceptScore W2542623543C61326573 @default.
- W2542623543 hasConceptScore W2542623543C62520636 @default.
- W2542623543 hasConceptScore W2542623543C76155785 @default.
- W2542623543 hasConceptScore W2542623543C94915269 @default.
- W2542623543 hasLocation W25426235431 @default.
- W2542623543 hasOpenAccess W2542623543 @default.
- W2542623543 hasPrimaryLocation W25426235431 @default.
- W2542623543 hasRelatedWork W1994129134 @default.
- W2542623543 hasRelatedWork W2026029549 @default.
- W2542623543 hasRelatedWork W2124086369 @default.
- W2542623543 hasRelatedWork W2128032727 @default.
- W2542623543 hasRelatedWork W2137576513 @default.
- W2542623543 hasRelatedWork W2224884787 @default.
- W2542623543 hasRelatedWork W2496034786 @default.
- W2542623543 hasRelatedWork W2755530029 @default.
- W2542623543 hasRelatedWork W2798234501 @default.
- W2542623543 hasRelatedWork W2952280410 @default.
- W2542623543 hasRelatedWork W2962828801 @default.
- W2542623543 hasRelatedWork W2963004024 @default.
- W2542623543 hasRelatedWork W2963122607 @default.
- W2542623543 hasRelatedWork W2963874901 @default.
- W2542623543 hasRelatedWork W2977844912 @default.
- W2542623543 hasRelatedWork W2991203304 @default.
- W2542623543 hasRelatedWork W3019156444 @default.
- W2542623543 hasRelatedWork W3093660365 @default.
- W2542623543 hasRelatedWork W3176523693 @default.
- W2542623543 hasRelatedWork W3195556713 @default.
- W2542623543 isParatext "false" @default.
- W2542623543 isRetracted "false" @default.
- W2542623543 magId "2542623543" @default.
- W2542623543 workType "article" @default.