Matches in SemOpenAlex for { <https://semopenalex.org/work/W2000200144> ?p ?o ?g. }
Showing items 1 to 86 of
86
with 100 items per page.
- W2000200144 abstract "Recent deep neural network systems for large vocabulary speech recognition are trained with minibatch stochastic gradient descent but use a variety of learning rate scheduling schemes. We investigate several of these schemes, particularly AdaGrad. Based on our analysis of its limitations, we propose a new variant `AdaDec' that decouples long-term learning-rate scheduling from per-parameter learning rate variation. AdaDec was found to result in higher frame accuracies than other methods. Overall, careful choice of learning rate schemes leads to faster convergence and lower word error rates." @default.
- W2000200144 created "2016-06-24" @default.
- W2000200144 creator A5031932067 @default.
- W2000200144 creator A5071956319 @default.
- W2000200144 creator A5077887756 @default.
- W2000200144 creator A5079708487 @default.
- W2000200144 date "2013-05-01" @default.
- W2000200144 modified "2023-10-18" @default.
- W2000200144 title "An empirical study of learning rates in deep neural networks for speech recognition" @default.
- W2000200144 cites W114517082 @default.
- W2000200144 cites W1994616650 @default.
- W2000200144 cites W2076794394 @default.
- W2000200144 cites W2086161653 @default.
- W2000200144 cites W2106051978 @default.
- W2000200144 cites W2130179707 @default.
- W2000200144 cites W2143908786 @default.
- W2000200144 cites W2160815625 @default.
- W2000200144 cites W2170541567 @default.
- W2000200144 cites W2296748324 @default.
- W2000200144 cites W2394932179 @default.
- W2000200144 cites W2403195671 @default.
- W2000200144 cites W4238404964 @default.
- W2000200144 doi "https://doi.org/10.1109/icassp.2013.6638963" @default.
- W2000200144 hasPublicationYear "2013" @default.
- W2000200144 type Work @default.
- W2000200144 sameAs 2000200144 @default.
- W2000200144 citedByCount "129" @default.
- W2000200144 countsByYear W20002001442013 @default.
- W2000200144 countsByYear W20002001442014 @default.
- W2000200144 countsByYear W20002001442015 @default.
- W2000200144 countsByYear W20002001442016 @default.
- W2000200144 countsByYear W20002001442017 @default.
- W2000200144 countsByYear W20002001442018 @default.
- W2000200144 countsByYear W20002001442019 @default.
- W2000200144 countsByYear W20002001442020 @default.
- W2000200144 countsByYear W20002001442021 @default.
- W2000200144 countsByYear W20002001442022 @default.
- W2000200144 countsByYear W20002001442023 @default.
- W2000200144 crossrefType "proceedings-article" @default.
- W2000200144 hasAuthorship W2000200144A5031932067 @default.
- W2000200144 hasAuthorship W2000200144A5071956319 @default.
- W2000200144 hasAuthorship W2000200144A5077887756 @default.
- W2000200144 hasAuthorship W2000200144A5079708487 @default.
- W2000200144 hasBestOaLocation W20002001442 @default.
- W2000200144 hasConcept C108583219 @default.
- W2000200144 hasConcept C119857082 @default.
- W2000200144 hasConcept C138885662 @default.
- W2000200144 hasConcept C153258448 @default.
- W2000200144 hasConcept C154945302 @default.
- W2000200144 hasConcept C206688291 @default.
- W2000200144 hasConcept C2777601683 @default.
- W2000200144 hasConcept C28490314 @default.
- W2000200144 hasConcept C40969351 @default.
- W2000200144 hasConcept C41008148 @default.
- W2000200144 hasConcept C41895202 @default.
- W2000200144 hasConcept C50644808 @default.
- W2000200144 hasConceptScore W2000200144C108583219 @default.
- W2000200144 hasConceptScore W2000200144C119857082 @default.
- W2000200144 hasConceptScore W2000200144C138885662 @default.
- W2000200144 hasConceptScore W2000200144C153258448 @default.
- W2000200144 hasConceptScore W2000200144C154945302 @default.
- W2000200144 hasConceptScore W2000200144C206688291 @default.
- W2000200144 hasConceptScore W2000200144C2777601683 @default.
- W2000200144 hasConceptScore W2000200144C28490314 @default.
- W2000200144 hasConceptScore W2000200144C40969351 @default.
- W2000200144 hasConceptScore W2000200144C41008148 @default.
- W2000200144 hasConceptScore W2000200144C41895202 @default.
- W2000200144 hasConceptScore W2000200144C50644808 @default.
- W2000200144 hasLocation W20002001441 @default.
- W2000200144 hasLocation W20002001442 @default.
- W2000200144 hasOpenAccess W2000200144 @default.
- W2000200144 hasPrimaryLocation W20002001441 @default.
- W2000200144 hasRelatedWork W2044529480 @default.
- W2000200144 hasRelatedWork W2792987183 @default.
- W2000200144 hasRelatedWork W3112316981 @default.
- W2000200144 hasRelatedWork W4223943233 @default.
- W2000200144 hasRelatedWork W4225161397 @default.
- W2000200144 hasRelatedWork W4248390167 @default.
- W2000200144 hasRelatedWork W4309045103 @default.
- W2000200144 hasRelatedWork W4312200629 @default.
- W2000200144 hasRelatedWork W4360585206 @default.
- W2000200144 hasRelatedWork W4364306694 @default.
- W2000200144 isParatext "false" @default.
- W2000200144 isRetracted "false" @default.
- W2000200144 magId "2000200144" @default.
- W2000200144 workType "article" @default.