Matches in SemOpenAlex for { <https://semopenalex.org/work/W3208712036> ?p ?o ?g. }
- W3208712036 abstract "We present a variational method for online state estimation and parameter learning in state-space models (SSMs), a ubiquitous class of latent variable models for sequential data. As per standard batch variational techniques, we use stochastic gradients to simultaneously optimize a lower bound on the log evidence with respect to both model parameters and a variational approximation of the states' posterior distribution. However, unlike existing approaches, our method is able to operate in an entirely online manner, such that historic observations do not require revisitation after being incorporated and the cost of updates at each time step remains constant, despite the growing dimensionality of the joint posterior distribution of the states. This is achieved by utilizing backward decompositions of this joint posterior distribution and of its variational approximation, combined with Bellman-type recursions for the evidence lower bound and its gradients. We demonstrate the performance of this methodology across several examples, including high-dimensional SSMs and sequential Variational Auto-Encoders." @default.
- W3208712036 created "2021-11-08" @default.
- W3208712036 creator A5044956245 @default.
- W3208712036 creator A5070107390 @default.
- W3208712036 creator A5078631467 @default.
- W3208712036 creator A5091677854 @default.
- W3208712036 date "2021-10-26" @default.
- W3208712036 modified "2023-09-27" @default.
- W3208712036 title "Online Variational Filtering and Parameter Learning" @default.
- W3208712036 cites W1522319528 @default.
- W3208712036 cites W1934021597 @default.
- W3208712036 cites W1959608418 @default.
- W3208712036 cites W2028907617 @default.
- W3208712036 cites W2061893297 @default.
- W3208712036 cites W2099371695 @default.
- W3208712036 cites W2121863487 @default.
- W3208712036 cites W2152767074 @default.
- W3208712036 cites W2169387407 @default.
- W3208712036 cites W2271522117 @default.
- W3208712036 cites W277831955 @default.
- W3208712036 cites W2799151646 @default.
- W3208712036 cites W2808697642 @default.
- W3208712036 cites W2891604023 @default.
- W3208712036 cites W2948627908 @default.
- W3208712036 cites W2963279312 @default.
- W3208712036 cites W2963319343 @default.
- W3208712036 cites W2963371290 @default.
- W3208712036 cites W2963556682 @default.
- W3208712036 cites W2963723789 @default.
- W3208712036 cites W2963775850 @default.
- W3208712036 cites W2964232608 @default.
- W3208712036 cites W2964334616 @default.
- W3208712036 cites W2970599228 @default.
- W3208712036 cites W2995546843 @default.
- W3208712036 cites W3015395304 @default.
- W3208712036 cites W3034297214 @default.
- W3208712036 cites W3035198488 @default.
- W3208712036 cites W3082258058 @default.
- W3208712036 cites W3103934441 @default.
- W3208712036 cites W3116168490 @default.
- W3208712036 cites W3170241221 @default.
- W3208712036 cites W3183169147 @default.
- W3208712036 cites W3201062265 @default.
- W3208712036 cites W592244745 @default.
- W3208712036 cites W68072195 @default.
- W3208712036 cites W88520345 @default.
- W3208712036 cites W2736839079 @default.
- W3208712036 hasPublicationYear "2021" @default.
- W3208712036 type Work @default.
- W3208712036 sameAs 3208712036 @default.
- W3208712036 citedByCount "0" @default.
- W3208712036 crossrefType "posted-content" @default.
- W3208712036 hasAuthorship W3208712036A5044956245 @default.
- W3208712036 hasAuthorship W3208712036A5070107390 @default.
- W3208712036 hasAuthorship W3208712036A5078631467 @default.
- W3208712036 hasAuthorship W3208712036A5091677854 @default.
- W3208712036 hasConcept C105795698 @default.
- W3208712036 hasConcept C107673813 @default.
- W3208712036 hasConcept C110121322 @default.
- W3208712036 hasConcept C111030470 @default.
- W3208712036 hasConcept C111919701 @default.
- W3208712036 hasConcept C11413529 @default.
- W3208712036 hasConcept C118505674 @default.
- W3208712036 hasConcept C126255220 @default.
- W3208712036 hasConcept C134306372 @default.
- W3208712036 hasConcept C154945302 @default.
- W3208712036 hasConcept C18653775 @default.
- W3208712036 hasConcept C199360897 @default.
- W3208712036 hasConcept C2777027219 @default.
- W3208712036 hasConcept C28826006 @default.
- W3208712036 hasConcept C33923547 @default.
- W3208712036 hasConcept C41008148 @default.
- W3208712036 hasConcept C51167844 @default.
- W3208712036 hasConcept C57830394 @default.
- W3208712036 hasConcept C64618476 @default.
- W3208712036 hasConcept C77553402 @default.
- W3208712036 hasConceptScore W3208712036C105795698 @default.
- W3208712036 hasConceptScore W3208712036C107673813 @default.
- W3208712036 hasConceptScore W3208712036C110121322 @default.
- W3208712036 hasConceptScore W3208712036C111030470 @default.
- W3208712036 hasConceptScore W3208712036C111919701 @default.
- W3208712036 hasConceptScore W3208712036C11413529 @default.
- W3208712036 hasConceptScore W3208712036C118505674 @default.
- W3208712036 hasConceptScore W3208712036C126255220 @default.
- W3208712036 hasConceptScore W3208712036C134306372 @default.
- W3208712036 hasConceptScore W3208712036C154945302 @default.
- W3208712036 hasConceptScore W3208712036C18653775 @default.
- W3208712036 hasConceptScore W3208712036C199360897 @default.
- W3208712036 hasConceptScore W3208712036C2777027219 @default.
- W3208712036 hasConceptScore W3208712036C28826006 @default.
- W3208712036 hasConceptScore W3208712036C33923547 @default.
- W3208712036 hasConceptScore W3208712036C41008148 @default.
- W3208712036 hasConceptScore W3208712036C51167844 @default.
- W3208712036 hasConceptScore W3208712036C57830394 @default.
- W3208712036 hasConceptScore W3208712036C64618476 @default.
- W3208712036 hasConceptScore W3208712036C77553402 @default.
- W3208712036 hasLocation W32087120361 @default.
- W3208712036 hasOpenAccess W3208712036 @default.
- W3208712036 hasPrimaryLocation W32087120361 @default.
- W3208712036 hasRelatedWork W1845773340 @default.