Matches in SemOpenAlex for { <https://semopenalex.org/work/W2938204024> ?p ?o ?g. }
- W2938204024 abstract "In this paper, we develop a stochastic-gradient learning algorithm for situations involving streaming data that arise from an underlying clustered structure. In such settings, the variance of gradient noise can be decomposed into the in-cluster variance $sigma _{{text{in}}}^2$ plus the between- cluster variance $sigma _{{text{bet}}}^2$. We develop a cluster-based online variance- reduced method (COVER) to eliminate $sigma _{{text{bet}}}^2$ and improve the MSD performance of stochastic-gradient descent (SGD) to the order of $Oleft( {sigma _{{text{in}}}^2} right)$. We establish the convergence property of COVER and derive a tight closed-form mean-square deviation (MSD) performance expression. Our simulations illustrate the improved performance of COVER in terms of steady-state performance." @default.
- W2938204024 created "2019-04-25" @default.
- W2938204024 creator A5005538082 @default.
- W2938204024 creator A5007519257 @default.
- W2938204024 creator A5091103447 @default.
- W2938204024 date "2019-05-01" @default.
- W2938204024 modified "2023-10-16" @default.
- W2938204024 title "COVER: A Cluster-based Variance Reduced Method for Online Learning" @default.
- W2938204024 cites W114517082 @default.
- W2938204024 cites W2012445782 @default.
- W2938204024 cites W2091825929 @default.
- W2938204024 cites W2107438106 @default.
- W2938204024 cites W2134603645 @default.
- W2938204024 cites W2135482703 @default.
- W2938204024 cites W2295500362 @default.
- W2938204024 cites W2527986773 @default.
- W2938204024 cites W2556219822 @default.
- W2938204024 cites W2610805269 @default.
- W2938204024 cites W2744692219 @default.
- W2938204024 cites W2963054052 @default.
- W2938204024 cites W2963156201 @default.
- W2938204024 cites W2964016923 @default.
- W2938204024 cites W2964112534 @default.
- W2938204024 cites W368469426 @default.
- W2938204024 doi "https://doi.org/10.1109/icassp.2019.8682527" @default.
- W2938204024 hasPublicationYear "2019" @default.
- W2938204024 type Work @default.
- W2938204024 sameAs 2938204024 @default.
- W2938204024 citedByCount "3" @default.
- W2938204024 countsByYear W29382040242020 @default.
- W2938204024 countsByYear W29382040242021 @default.
- W2938204024 crossrefType "proceedings-article" @default.
- W2938204024 hasAuthorship W2938204024A5005538082 @default.
- W2938204024 hasAuthorship W2938204024A5007519257 @default.
- W2938204024 hasAuthorship W2938204024A5091103447 @default.
- W2938204024 hasConcept C11413529 @default.
- W2938204024 hasConcept C115961682 @default.
- W2938204024 hasConcept C121332964 @default.
- W2938204024 hasConcept C121955636 @default.
- W2938204024 hasConcept C127413603 @default.
- W2938204024 hasConcept C144133560 @default.
- W2938204024 hasConcept C154945302 @default.
- W2938204024 hasConcept C162324750 @default.
- W2938204024 hasConcept C164866538 @default.
- W2938204024 hasConcept C196083921 @default.
- W2938204024 hasConcept C199360897 @default.
- W2938204024 hasConcept C206688291 @default.
- W2938204024 hasConcept C2777303404 @default.
- W2938204024 hasConcept C2778049214 @default.
- W2938204024 hasConcept C2780428219 @default.
- W2938204024 hasConcept C33923547 @default.
- W2938204024 hasConcept C41008148 @default.
- W2938204024 hasConcept C50522688 @default.
- W2938204024 hasConcept C50644808 @default.
- W2938204024 hasConcept C62520636 @default.
- W2938204024 hasConcept C78519656 @default.
- W2938204024 hasConcept C99498987 @default.
- W2938204024 hasConceptScore W2938204024C11413529 @default.
- W2938204024 hasConceptScore W2938204024C115961682 @default.
- W2938204024 hasConceptScore W2938204024C121332964 @default.
- W2938204024 hasConceptScore W2938204024C121955636 @default.
- W2938204024 hasConceptScore W2938204024C127413603 @default.
- W2938204024 hasConceptScore W2938204024C144133560 @default.
- W2938204024 hasConceptScore W2938204024C154945302 @default.
- W2938204024 hasConceptScore W2938204024C162324750 @default.
- W2938204024 hasConceptScore W2938204024C164866538 @default.
- W2938204024 hasConceptScore W2938204024C196083921 @default.
- W2938204024 hasConceptScore W2938204024C199360897 @default.
- W2938204024 hasConceptScore W2938204024C206688291 @default.
- W2938204024 hasConceptScore W2938204024C2777303404 @default.
- W2938204024 hasConceptScore W2938204024C2778049214 @default.
- W2938204024 hasConceptScore W2938204024C2780428219 @default.
- W2938204024 hasConceptScore W2938204024C33923547 @default.
- W2938204024 hasConceptScore W2938204024C41008148 @default.
- W2938204024 hasConceptScore W2938204024C50522688 @default.
- W2938204024 hasConceptScore W2938204024C50644808 @default.
- W2938204024 hasConceptScore W2938204024C62520636 @default.
- W2938204024 hasConceptScore W2938204024C78519656 @default.
- W2938204024 hasConceptScore W2938204024C99498987 @default.
- W2938204024 hasLocation W29382040241 @default.
- W2938204024 hasOpenAccess W2938204024 @default.
- W2938204024 hasPrimaryLocation W29382040241 @default.
- W2938204024 hasRelatedWork W1565442263 @default.
- W2938204024 hasRelatedWork W2002870710 @default.
- W2938204024 hasRelatedWork W2019022041 @default.
- W2938204024 hasRelatedWork W2097560589 @default.
- W2938204024 hasRelatedWork W2116293245 @default.
- W2938204024 hasRelatedWork W2139155312 @default.
- W2938204024 hasRelatedWork W2172057042 @default.
- W2938204024 hasRelatedWork W2267151711 @default.
- W2938204024 hasRelatedWork W2535933102 @default.
- W2938204024 hasRelatedWork W2537323706 @default.
- W2938204024 hasRelatedWork W2545080023 @default.
- W2938204024 hasRelatedWork W2962901793 @default.
- W2938204024 hasRelatedWork W2981631696 @default.
- W2938204024 hasRelatedWork W2988903469 @default.
- W2938204024 hasRelatedWork W2998409652 @default.
- W2938204024 hasRelatedWork W307213043 @default.
- W2938204024 hasRelatedWork W3092183261 @default.
- W2938204024 hasRelatedWork W3150749950 @default.