Matches in SemOpenAlex for { <https://semopenalex.org/work/W2949910669> ?p ?o ?g. }
- W2949910669 abstract "Deep Gaussian processes (DGPs) can model complex marginal densities as well as complex mappings. Non-Gaussian marginals are essential for modelling real-world data, and can be generated from the DGP by incorporating uncorrelated variables to the model. Previous work on DGP models has introduced noise additively and used variational inference with a combination of sparse Gaussian processes and mean-field Gaussians for the approximate posterior. Additive noise attenuates the signal, and the Gaussian form of variational distribution may lead to an inaccurate posterior. We instead incorporate noisy variables as latent covariates, and propose a novel importance-weighted objective, which leverages analytic results and provides a mechanism to trade off computation for improved accuracy. Our results demonstrate that the importance-weighted objective works well in practice and consistently outperforms classical variational inference, especially for deeper models." @default.
- W2949910669 created "2019-06-27" @default.
- W2949910669 creator A5001763022 @default.
- W2949910669 creator A5023796924 @default.
- W2949910669 creator A5085437275 @default.
- W2949910669 creator A5090767252 @default.
- W2949910669 date "2019-05-14" @default.
- W2949910669 modified "2023-09-27" @default.
- W2949910669 title "Deep Gaussian Processes with Importance-Weighted Variational Inference" @default.
- W2949910669 cites W137285897 @default.
- W2949910669 cites W1533861849 @default.
- W2949910669 cites W1589901016 @default.
- W2949910669 cites W1641973859 @default.
- W2949910669 cites W1777124189 @default.
- W2949910669 cites W1830850394 @default.
- W2949910669 cites W1950803081 @default.
- W2949910669 cites W1959608418 @default.
- W2949910669 cites W196162695 @default.
- W2949910669 cites W2138806252 @default.
- W2949910669 cites W2146611938 @default.
- W2949910669 cites W215707797 @default.
- W2949910669 cites W2169779569 @default.
- W2949910669 cites W2188365844 @default.
- W2949910669 cites W2194775991 @default.
- W2949910669 cites W2257113116 @default.
- W2949910669 cites W2540189295 @default.
- W2949910669 cites W2559331598 @default.
- W2949910669 cites W2737610083 @default.
- W2949910669 cites W2889028244 @default.
- W2949910669 cites W2944370630 @default.
- W2949910669 cites W2953263857 @default.
- W2949910669 cites W2962779017 @default.
- W2949910669 cites W2962875063 @default.
- W2949910669 cites W2962897886 @default.
- W2949910669 cites W2962947707 @default.
- W2949910669 cites W2963192657 @default.
- W2949910669 cites W2963275229 @default.
- W2949910669 cites W2963446624 @default.
- W2949910669 cites W2963711523 @default.
- W2949910669 cites W2963798377 @default.
- W2949910669 cites W2963827215 @default.
- W2949910669 cites W2963935178 @default.
- W2949910669 cites W2964121744 @default.
- W2949910669 cites W2964141597 @default.
- W2949910669 cites W2964267629 @default.
- W2949910669 cites W66306528 @default.
- W2949910669 hasPublicationYear "2019" @default.
- W2949910669 type Work @default.
- W2949910669 sameAs 2949910669 @default.
- W2949910669 citedByCount "0" @default.
- W2949910669 crossrefType "posted-content" @default.
- W2949910669 hasAuthorship W2949910669A5001763022 @default.
- W2949910669 hasAuthorship W2949910669A5023796924 @default.
- W2949910669 hasAuthorship W2949910669A5085437275 @default.
- W2949910669 hasAuthorship W2949910669A5090767252 @default.
- W2949910669 hasConcept C11413529 @default.
- W2949910669 hasConcept C115961682 @default.
- W2949910669 hasConcept C121332964 @default.
- W2949910669 hasConcept C154945302 @default.
- W2949910669 hasConcept C163716315 @default.
- W2949910669 hasConcept C2776214188 @default.
- W2949910669 hasConcept C28826006 @default.
- W2949910669 hasConcept C33923547 @default.
- W2949910669 hasConcept C41008148 @default.
- W2949910669 hasConcept C4199805 @default.
- W2949910669 hasConcept C51267290 @default.
- W2949910669 hasConcept C61326573 @default.
- W2949910669 hasConcept C62520636 @default.
- W2949910669 hasConcept C99498987 @default.
- W2949910669 hasConceptScore W2949910669C11413529 @default.
- W2949910669 hasConceptScore W2949910669C115961682 @default.
- W2949910669 hasConceptScore W2949910669C121332964 @default.
- W2949910669 hasConceptScore W2949910669C154945302 @default.
- W2949910669 hasConceptScore W2949910669C163716315 @default.
- W2949910669 hasConceptScore W2949910669C2776214188 @default.
- W2949910669 hasConceptScore W2949910669C28826006 @default.
- W2949910669 hasConceptScore W2949910669C33923547 @default.
- W2949910669 hasConceptScore W2949910669C41008148 @default.
- W2949910669 hasConceptScore W2949910669C4199805 @default.
- W2949910669 hasConceptScore W2949910669C51267290 @default.
- W2949910669 hasConceptScore W2949910669C61326573 @default.
- W2949910669 hasConceptScore W2949910669C62520636 @default.
- W2949910669 hasConceptScore W2949910669C99498987 @default.
- W2949910669 hasLocation W29499106691 @default.
- W2949910669 hasOpenAccess W2949910669 @default.
- W2949910669 hasPrimaryLocation W29499106691 @default.
- W2949910669 hasRelatedWork W1513910483 @default.
- W2949910669 hasRelatedWork W1762233554 @default.
- W2949910669 hasRelatedWork W1897171481 @default.
- W2949910669 hasRelatedWork W2132849848 @default.
- W2949910669 hasRelatedWork W2395877431 @default.
- W2949910669 hasRelatedWork W2539640139 @default.
- W2949910669 hasRelatedWork W2540539267 @default.
- W2949910669 hasRelatedWork W2735731743 @default.
- W2949910669 hasRelatedWork W2766007589 @default.
- W2949910669 hasRelatedWork W2924342202 @default.
- W2949910669 hasRelatedWork W2949546462 @default.
- W2949910669 hasRelatedWork W2963717915 @default.
- W2949910669 hasRelatedWork W2982247681 @default.
- W2949910669 hasRelatedWork W2993626171 @default.