Matches in SemOpenAlex for { <https://semopenalex.org/work/W2103369448> ?p ?o ?g. }
Showing items 1 to 90 of
90
with 100 items per page.
- W2103369448 endingPage "1544" @default.
- W2103369448 startingPage "1507" @default.
- W2103369448 abstract "We present a Monte Carlo approach for training partially observable diffusion processes. We apply the approach to diffusion networks, a stochastic version of continuous recurrent neural networks. The approach is aimed at learning probability distributions of continuous paths, not just expected values. Interestingly, the relevant activation statistics used by the learning rule presented here are inner products in the Hilbert space of square integrable functions. These inner products can be computed using Hebbian operations and do not require backpropagation of error signals. Moreover, standard kernel methods could potentially be applied to compute such inner products. We propose that the main reason that recurrent neural networks have not worked well in engineering applications (e.g., speech recognition) is that they implicitly rely on a very simplistic likelihood model. The diffusion network approach proposed here is much richer and may open new avenues for applications of recurrent neural networks. We present some analysis and simulations to support this view. Very encouraging results were obtained on a visual speech recognition task in which neural networks outperformed hidden Markov models." @default.
- W2103369448 created "2016-06-24" @default.
- W2103369448 creator A5006692623 @default.
- W2103369448 creator A5009698276 @default.
- W2103369448 creator A5010136172 @default.
- W2103369448 date "2002-07-01" @default.
- W2103369448 modified "2023-09-24" @default.
- W2103369448 title "A Monte Carlo EM Approach for Partially Observable Diffusion Processes: Theory and Applications to Neural Networks" @default.
- W2103369448 cites W1963718895 @default.
- W2103369448 cites W2005573962 @default.
- W2103369448 cites W2042492924 @default.
- W2103369448 cites W2067773920 @default.
- W2103369448 cites W2083402998 @default.
- W2103369448 cites W2087348129 @default.
- W2103369448 cites W2091617878 @default.
- W2103369448 cites W2103030897 @default.
- W2103369448 cites W2112246162 @default.
- W2103369448 cites W2129618392 @default.
- W2103369448 cites W2137983211 @default.
- W2103369448 cites W2139212933 @default.
- W2103369448 cites W2154890045 @default.
- W2103369448 cites W4240657734 @default.
- W2103369448 doi "https://doi.org/10.1162/08997660260028593" @default.
- W2103369448 hasPubMedId "https://pubmed.ncbi.nlm.nih.gov/12079544" @default.
- W2103369448 hasPublicationYear "2002" @default.
- W2103369448 type Work @default.
- W2103369448 sameAs 2103369448 @default.
- W2103369448 citedByCount "25" @default.
- W2103369448 countsByYear W21033694482012 @default.
- W2103369448 countsByYear W21033694482014 @default.
- W2103369448 countsByYear W21033694482015 @default.
- W2103369448 countsByYear W21033694482017 @default.
- W2103369448 countsByYear W21033694482019 @default.
- W2103369448 countsByYear W21033694482020 @default.
- W2103369448 countsByYear W21033694482021 @default.
- W2103369448 countsByYear W21033694482023 @default.
- W2103369448 crossrefType "journal-article" @default.
- W2103369448 hasAuthorship W2103369448A5006692623 @default.
- W2103369448 hasAuthorship W2103369448A5009698276 @default.
- W2103369448 hasAuthorship W2103369448A5010136172 @default.
- W2103369448 hasConcept C105795698 @default.
- W2103369448 hasConcept C11413529 @default.
- W2103369448 hasConcept C114614502 @default.
- W2103369448 hasConcept C119857082 @default.
- W2103369448 hasConcept C147168706 @default.
- W2103369448 hasConcept C154945302 @default.
- W2103369448 hasConcept C155032097 @default.
- W2103369448 hasConcept C192576344 @default.
- W2103369448 hasConcept C19499675 @default.
- W2103369448 hasConcept C23224414 @default.
- W2103369448 hasConcept C33923547 @default.
- W2103369448 hasConcept C41008148 @default.
- W2103369448 hasConcept C50644808 @default.
- W2103369448 hasConcept C74193536 @default.
- W2103369448 hasConceptScore W2103369448C105795698 @default.
- W2103369448 hasConceptScore W2103369448C11413529 @default.
- W2103369448 hasConceptScore W2103369448C114614502 @default.
- W2103369448 hasConceptScore W2103369448C119857082 @default.
- W2103369448 hasConceptScore W2103369448C147168706 @default.
- W2103369448 hasConceptScore W2103369448C154945302 @default.
- W2103369448 hasConceptScore W2103369448C155032097 @default.
- W2103369448 hasConceptScore W2103369448C192576344 @default.
- W2103369448 hasConceptScore W2103369448C19499675 @default.
- W2103369448 hasConceptScore W2103369448C23224414 @default.
- W2103369448 hasConceptScore W2103369448C33923547 @default.
- W2103369448 hasConceptScore W2103369448C41008148 @default.
- W2103369448 hasConceptScore W2103369448C50644808 @default.
- W2103369448 hasConceptScore W2103369448C74193536 @default.
- W2103369448 hasIssue "7" @default.
- W2103369448 hasLocation W21033694481 @default.
- W2103369448 hasLocation W21033694482 @default.
- W2103369448 hasOpenAccess W2103369448 @default.
- W2103369448 hasPrimaryLocation W21033694481 @default.
- W2103369448 hasRelatedWork W1604061580 @default.
- W2103369448 hasRelatedWork W1814702984 @default.
- W2103369448 hasRelatedWork W2048308819 @default.
- W2103369448 hasRelatedWork W2362189222 @default.
- W2103369448 hasRelatedWork W2391384657 @default.
- W2103369448 hasRelatedWork W3044458868 @default.
- W2103369448 hasRelatedWork W4317242789 @default.
- W2103369448 hasRelatedWork W1629725936 @default.
- W2103369448 hasRelatedWork W2182757990 @default.
- W2103369448 hasRelatedWork W2516580779 @default.
- W2103369448 hasVolume "14" @default.
- W2103369448 isParatext "false" @default.
- W2103369448 isRetracted "false" @default.
- W2103369448 magId "2103369448" @default.
- W2103369448 workType "article" @default.