Matches in SemOpenAlex for { <https://semopenalex.org/work/W3114193967> ?p ?o ?g. }
Showing items 1 to 97 of
97
with 100 items per page.
- W3114193967 abstract "We study batch normalisation in the context of variational inference methods in Bayesian neural networks, such as mean-field or MC Dropout. We show that batch-normalisation does not affect the optimum of the evidence lower bound (ELBO). Furthermore, we study the Monte Carlo Batch Normalisation (MCBN) algorithm, proposed as an approximate inference technique parallel to MC Dropout, and show that for larger batch sizes, MCBN fails to capture epistemic uncertainty. Finally, we provide insights into what is required to fix this failure, namely having to view the mini-batch size as a variational parameter in MCBN. We comment on the asymptotics of the ELBO with respect to this variational parameter, showing that as dataset size increases towards infinity, the batch-size must increase towards infinity as well for MCBN to be a valid approximate inference technique." @default.
- W3114193967 created "2021-01-05" @default.
- W3114193967 creator A5017379207 @default.
- W3114193967 creator A5029186201 @default.
- W3114193967 creator A5042899882 @default.
- W3114193967 creator A5043440142 @default.
- W3114193967 date "2020-12-24" @default.
- W3114193967 modified "2023-09-27" @default.
- W3114193967 title "On Batch Normalisation for Approximate Bayesian Inference." @default.
- W3114193967 cites W1719489212 @default.
- W3114193967 cites W2047229728 @default.
- W3114193967 cites W2108677974 @default.
- W3114193967 cites W2117670920 @default.
- W3114193967 cites W2139137304 @default.
- W3114193967 cites W2166851633 @default.
- W3114193967 cites W2335728318 @default.
- W3114193967 cites W2581082771 @default.
- W3114193967 cites W2600383743 @default.
- W3114193967 cites W2750384547 @default.
- W3114193967 cites W2752013927 @default.
- W3114193967 cites W2901746906 @default.
- W3114193967 cites W2914484425 @default.
- W3114193967 cites W2949117887 @default.
- W3114193967 cites W2949738940 @default.
- W3114193967 cites W2951266961 @default.
- W3114193967 cites W2964059111 @default.
- W3114193967 cites W3118608800 @default.
- W3114193967 hasPublicationYear "2020" @default.
- W3114193967 type Work @default.
- W3114193967 sameAs 3114193967 @default.
- W3114193967 citedByCount "0" @default.
- W3114193967 crossrefType "posted-content" @default.
- W3114193967 hasAuthorship W3114193967A5017379207 @default.
- W3114193967 hasAuthorship W3114193967A5029186201 @default.
- W3114193967 hasAuthorship W3114193967A5042899882 @default.
- W3114193967 hasAuthorship W3114193967A5043440142 @default.
- W3114193967 hasConcept C105795698 @default.
- W3114193967 hasConcept C107673813 @default.
- W3114193967 hasConcept C11413529 @default.
- W3114193967 hasConcept C119857082 @default.
- W3114193967 hasConcept C134306372 @default.
- W3114193967 hasConcept C151730666 @default.
- W3114193967 hasConcept C154945302 @default.
- W3114193967 hasConcept C160234255 @default.
- W3114193967 hasConcept C19499675 @default.
- W3114193967 hasConcept C2776145597 @default.
- W3114193967 hasConcept C2776214188 @default.
- W3114193967 hasConcept C2779343474 @default.
- W3114193967 hasConcept C33923547 @default.
- W3114193967 hasConcept C41008148 @default.
- W3114193967 hasConcept C7321624 @default.
- W3114193967 hasConcept C77553402 @default.
- W3114193967 hasConcept C86803240 @default.
- W3114193967 hasConceptScore W3114193967C105795698 @default.
- W3114193967 hasConceptScore W3114193967C107673813 @default.
- W3114193967 hasConceptScore W3114193967C11413529 @default.
- W3114193967 hasConceptScore W3114193967C119857082 @default.
- W3114193967 hasConceptScore W3114193967C134306372 @default.
- W3114193967 hasConceptScore W3114193967C151730666 @default.
- W3114193967 hasConceptScore W3114193967C154945302 @default.
- W3114193967 hasConceptScore W3114193967C160234255 @default.
- W3114193967 hasConceptScore W3114193967C19499675 @default.
- W3114193967 hasConceptScore W3114193967C2776145597 @default.
- W3114193967 hasConceptScore W3114193967C2776214188 @default.
- W3114193967 hasConceptScore W3114193967C2779343474 @default.
- W3114193967 hasConceptScore W3114193967C33923547 @default.
- W3114193967 hasConceptScore W3114193967C41008148 @default.
- W3114193967 hasConceptScore W3114193967C7321624 @default.
- W3114193967 hasConceptScore W3114193967C77553402 @default.
- W3114193967 hasConceptScore W3114193967C86803240 @default.
- W3114193967 hasLocation W31141939671 @default.
- W3114193967 hasOpenAccess W3114193967 @default.
- W3114193967 hasPrimaryLocation W31141939671 @default.
- W3114193967 hasRelatedWork W1964226618 @default.
- W3114193967 hasRelatedWork W1985339837 @default.
- W3114193967 hasRelatedWork W2136097307 @default.
- W3114193967 hasRelatedWork W2535941552 @default.
- W3114193967 hasRelatedWork W2607585387 @default.
- W3114193967 hasRelatedWork W2739217148 @default.
- W3114193967 hasRelatedWork W2805502496 @default.
- W3114193967 hasRelatedWork W2912819213 @default.
- W3114193967 hasRelatedWork W2949651817 @default.
- W3114193967 hasRelatedWork W2950883651 @default.
- W3114193967 hasRelatedWork W2951222995 @default.
- W3114193967 hasRelatedWork W2967542826 @default.
- W3114193967 hasRelatedWork W2971403200 @default.
- W3114193967 hasRelatedWork W2990573086 @default.
- W3114193967 hasRelatedWork W3091291905 @default.
- W3114193967 hasRelatedWork W3093782733 @default.
- W3114193967 hasRelatedWork W3101747691 @default.
- W3114193967 hasRelatedWork W3105385755 @default.
- W3114193967 hasRelatedWork W3121853896 @default.
- W3114193967 hasRelatedWork W599376124 @default.
- W3114193967 isParatext "false" @default.
- W3114193967 isRetracted "false" @default.
- W3114193967 magId "3114193967" @default.
- W3114193967 workType "article" @default.