FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W2769603572> ?p ?o ?g. }

• W2769603572 abstract "We present a general-purpose method to train Markov chain Monte Carlo kernels, parameterized by deep neural networks, that converge and mix quickly to their target distribution. Our method generalizes Hamiltonian Monte Carlo and is trained to maximize expected squared jumped distance, a proxy for mixing speed. We demonstrate large empirical gains on a collection of simple but challenging distributions, for instance achieving a 106x improvement in effective sample size in one case, and mixing when standard HMC makes no measurable progress in a second. Finally, we show quantitative and qualitative gains on a real-world task: latent-variable generative modeling. We release an open source TensorFlow implementation of the algorithm." @default.
• W2769603572 created "2017-12-04" @default.
• W2769603572 creator A5064986566 @default.
• W2769603572 creator A5069219684 @default.
• W2769603572 creator A5086319422 @default.
• W2769603572 date "2017-11-25" @default.
• W2769603572 modified "2023-09-27" @default.
• W2769603572 title "Generalizing Hamiltonian Monte Carlo with Neural Networks" @default.
• W2769603572 cites W1513873506 @default.
• W2769603572 cites W1522301498 @default.
• W2769603572 cites W1655728402 @default.
• W2769603572 cites W1705521864 @default.
• W2769603572 cites W1826892964 @default.
• W2769603572 cites W1906598733 @default.
• W2769603572 cites W195465510 @default.
• W2769603572 cites W1982032572 @default.
• W2769603572 cites W1995780830 @default.
• W2769603572 cites W2038334672 @default.
• W2769603572 cites W2043719428 @default.
• W2769603572 cites W2046290306 @default.
• W2769603572 cites W2058763860 @default.
• W2769603572 cites W2099057450 @default.
• W2769603572 cites W2125155341 @default.
• W2769603572 cites W2126398289 @default.
• W2769603572 cites W2129069237 @default.
• W2769603572 cites W2132535396 @default.
• W2769603572 cites W2134949445 @default.
• W2769603572 cites W2135194391 @default.
• W2769603572 cites W2135973421 @default.
• W2769603572 cites W2138309709 @default.
• W2769603572 cites W2142274086 @default.
• W2769603572 cites W2160065175 @default.
• W2769603572 cites W2204383650 @default.
• W2769603572 cites W2259511957 @default.
• W2769603572 cites W2409550820 @default.
• W2769603572 cites W2478027467 @default.
• W2769603572 cites W2506672536 @default.
• W2769603572 cites W2530117613 @default.
• W2769603572 cites W2592477384 @default.
• W2769603572 cites W2739517340 @default.
• W2769603572 cites W2951004968 @default.
• W2769603572 cites W2951105989 @default.
• W2769603572 cites W2953116488 @default.
• W2769603572 cites W2963565380 @default.
• W2769603572 cites W2963977107 @default.
• W2769603572 cites W3140968660 @default.
• W2769603572 hasPublicationYear "2017" @default.
• W2769603572 type Work @default.
• W2769603572 sameAs 2769603572 @default.
• W2769603572 citedByCount "10" @default.
• W2769603572 countsByYear W27696035722018 @default.
• W2769603572 countsByYear W27696035722019 @default.
• W2769603572 countsByYear W27696035722020 @default.
• W2769603572 crossrefType "posted-content" @default.
• W2769603572 hasAuthorship W2769603572A5064986566 @default.
• W2769603572 hasAuthorship W2769603572A5069219684 @default.
• W2769603572 hasAuthorship W2769603572A5086319422 @default.
• W2769603572 hasConcept C105795698 @default.
• W2769603572 hasConcept C107673813 @default.
• W2769603572 hasConcept C111350023 @default.
• W2769603572 hasConcept C11413529 @default.
• W2769603572 hasConcept C119857082 @default.
• W2769603572 hasConcept C126255220 @default.
• W2769603572 hasConcept C130787639 @default.
• W2769603572 hasConcept C13153151 @default.
• W2769603572 hasConcept C154945302 @default.
• W2769603572 hasConcept C165464430 @default.
• W2769603572 hasConcept C19499675 @default.
• W2769603572 hasConcept C28826006 @default.
• W2769603572 hasConcept C33923547 @default.
• W2769603572 hasConcept C41008148 @default.
• W2769603572 hasConcept C50644808 @default.
• W2769603572 hasConcept C51167844 @default.
• W2769603572 hasConcept C98763669 @default.
• W2769603572 hasConceptScore W2769603572C105795698 @default.
• W2769603572 hasConceptScore W2769603572C107673813 @default.
• W2769603572 hasConceptScore W2769603572C111350023 @default.
• W2769603572 hasConceptScore W2769603572C11413529 @default.
• W2769603572 hasConceptScore W2769603572C119857082 @default.
• W2769603572 hasConceptScore W2769603572C126255220 @default.
• W2769603572 hasConceptScore W2769603572C130787639 @default.
• W2769603572 hasConceptScore W2769603572C13153151 @default.
• W2769603572 hasConceptScore W2769603572C154945302 @default.
• W2769603572 hasConceptScore W2769603572C165464430 @default.
• W2769603572 hasConceptScore W2769603572C19499675 @default.
• W2769603572 hasConceptScore W2769603572C28826006 @default.
• W2769603572 hasConceptScore W2769603572C33923547 @default.
• W2769603572 hasConceptScore W2769603572C41008148 @default.
• W2769603572 hasConceptScore W2769603572C50644808 @default.
• W2769603572 hasConceptScore W2769603572C51167844 @default.
• W2769603572 hasConceptScore W2769603572C98763669 @default.
• W2769603572 hasLocation W27696035721 @default.
• W2769603572 hasOpenAccess W2769603572 @default.
• W2769603572 hasPrimaryLocation W27696035721 @default.
• W2769603572 hasRelatedWork W1522301498 @default.
• W2769603572 hasRelatedWork W2122828856 @default.
• W2769603572 hasRelatedWork W2144193737 @default.
• W2769603572 hasRelatedWork W2346910172 @default.
• W2769603572 hasRelatedWork W2409550820 @default.
• W2769603572 hasRelatedWork W2574834547 @default.

• next