Matches in SemOpenAlex for { <https://semopenalex.org/work/W2763183857> ?p ?o ?g. }
- W2763183857 abstract "A key task in Bayesian statistics is sampling from distributions that are only specified up to a partition function (i.e., constant of proportionality). However, without any assumptions, sampling (even approximately) can be #P-hard, and few works have provided beyond worst-case guarantees for such settings. For log-concave distributions, classical results going back to Bakry and Emery (1985) show that natural continuous-time Markov chains called Langevin diffusions mix in polynomial time. The most salient feature of log-concavity violated in practice is uni-modality: commonly, the distributions we wish to sample from are multi-modal. In the presence of multiple deep and well-separated modes, Langevin diffusion suffers from torpid mixing. We address this problem by combining Langevin diffusion with simulated tempering. The result is a Markov chain that mixes more rapidly by transitioning between different temperatures of the distribution. We analyze this Markov chain for the canonical multi-modal distribution: a mixture of gaussians (of equal variance). The algorithm based on our Markov chain provably samples from distributions that are close to mixtures of gaussians, given access to the gradient of the log-pdf. For the analysis, we use a spectral decomposition theorem for graphs (Gharan and Trevisan, 2014) and a Markov chain decomposition technique (Madras and Randall, 2002)." @default.
- W2763183857 created "2017-10-20" @default.
- W2763183857 creator A5008139378 @default.
- W2763183857 creator A5014972240 @default.
- W2763183857 creator A5035001911 @default.
- W2763183857 date "2017-10-07" @default.
- W2763183857 modified "2023-09-27" @default.
- W2763183857 title "Beyond Log-concavity: Provable Guarantees for Sampling Multi-modal Distributions using Simulated Tempering Langevin Monte Carlo" @default.
- W2763183857 cites W1560153690 @default.
- W2763183857 cites W1909320841 @default.
- W2763183857 cites W1938536253 @default.
- W2763183857 cites W1976418263 @default.
- W2763183857 cites W1976706540 @default.
- W2763183857 cites W2001464954 @default.
- W2763183857 cites W2003268963 @default.
- W2763183857 cites W2014565165 @default.
- W2763183857 cites W2040060987 @default.
- W2763183857 cites W2046392353 @default.
- W2763183857 cites W2090987881 @default.
- W2763183857 cites W2105459687 @default.
- W2763183857 cites W2110435328 @default.
- W2763183857 cites W2143238292 @default.
- W2763183857 cites W2161269121 @default.
- W2763183857 cites W2222154095 @default.
- W2763183857 cites W2562674776 @default.
- W2763183857 cites W2565444551 @default.
- W2763183857 cites W2590513847 @default.
- W2763183857 cites W2951004968 @default.
- W2763183857 cites W2952389066 @default.
- W2763183857 cites W2953337630 @default.
- W2763183857 cites W2963625764 @default.
- W2763183857 cites W2963627928 @default.
- W2763183857 cites W2963870478 @default.
- W2763183857 cites W423603040 @default.
- W2763183857 cites W825717146 @default.
- W2763183857 hasPublicationYear "2017" @default.
- W2763183857 type Work @default.
- W2763183857 sameAs 2763183857 @default.
- W2763183857 citedByCount "5" @default.
- W2763183857 countsByYear W27631838572017 @default.
- W2763183857 countsByYear W27631838572018 @default.
- W2763183857 countsByYear W27631838572020 @default.
- W2763183857 crossrefType "posted-content" @default.
- W2763183857 hasAuthorship W2763183857A5008139378 @default.
- W2763183857 hasAuthorship W2763183857A5014972240 @default.
- W2763183857 hasAuthorship W2763183857A5035001911 @default.
- W2763183857 hasConcept C105795698 @default.
- W2763183857 hasConcept C111350023 @default.
- W2763183857 hasConcept C11413529 @default.
- W2763183857 hasConcept C121332964 @default.
- W2763183857 hasConcept C121864883 @default.
- W2763183857 hasConcept C163836022 @default.
- W2763183857 hasConcept C189973286 @default.
- W2763183857 hasConcept C19499675 @default.
- W2763183857 hasConcept C2778401447 @default.
- W2763183857 hasConcept C28826006 @default.
- W2763183857 hasConcept C33923547 @default.
- W2763183857 hasConcept C62520636 @default.
- W2763183857 hasConcept C97074811 @default.
- W2763183857 hasConcept C98763669 @default.
- W2763183857 hasConceptScore W2763183857C105795698 @default.
- W2763183857 hasConceptScore W2763183857C111350023 @default.
- W2763183857 hasConceptScore W2763183857C11413529 @default.
- W2763183857 hasConceptScore W2763183857C121332964 @default.
- W2763183857 hasConceptScore W2763183857C121864883 @default.
- W2763183857 hasConceptScore W2763183857C163836022 @default.
- W2763183857 hasConceptScore W2763183857C189973286 @default.
- W2763183857 hasConceptScore W2763183857C19499675 @default.
- W2763183857 hasConceptScore W2763183857C2778401447 @default.
- W2763183857 hasConceptScore W2763183857C28826006 @default.
- W2763183857 hasConceptScore W2763183857C33923547 @default.
- W2763183857 hasConceptScore W2763183857C62520636 @default.
- W2763183857 hasConceptScore W2763183857C97074811 @default.
- W2763183857 hasConceptScore W2763183857C98763669 @default.
- W2763183857 hasLocation W27631838571 @default.
- W2763183857 hasOpenAccess W2763183857 @default.
- W2763183857 hasPrimaryLocation W27631838571 @default.
- W2763183857 hasRelatedWork W1519211266 @default.
- W2763183857 hasRelatedWork W1544503773 @default.
- W2763183857 hasRelatedWork W1600627921 @default.
- W2763183857 hasRelatedWork W1855009442 @default.
- W2763183857 hasRelatedWork W1965828084 @default.
- W2763183857 hasRelatedWork W1983628095 @default.
- W2763183857 hasRelatedWork W2011040751 @default.
- W2763183857 hasRelatedWork W2052660758 @default.
- W2763183857 hasRelatedWork W2068927766 @default.
- W2763183857 hasRelatedWork W2077492312 @default.
- W2763183857 hasRelatedWork W2196919871 @default.
- W2763183857 hasRelatedWork W2554560886 @default.
- W2763183857 hasRelatedWork W2751244547 @default.
- W2763183857 hasRelatedWork W2902106771 @default.
- W2763183857 hasRelatedWork W2924841542 @default.
- W2763183857 hasRelatedWork W2951363171 @default.
- W2763183857 hasRelatedWork W2963858858 @default.
- W2763183857 hasRelatedWork W2972743405 @default.
- W2763183857 hasRelatedWork W887548946 @default.
- W2763183857 hasRelatedWork W91361971 @default.
- W2763183857 isParatext "false" @default.
- W2763183857 isRetracted "false" @default.
- W2763183857 magId "2763183857" @default.