Matches in SemOpenAlex for { <https://semopenalex.org/work/W2979789219> ?p ?o ?g. }
- W2979789219 abstract "We introduce a pruning algorithm that provably sparsifies the parameters of a trained model in a way that approximately preserves the model's predictive accuracy. Our algorithm uses a small batch of input points to construct a data-informed importance sampling distribution over the network's parameters, and adaptively mixes a sampling-based and deterministic pruning procedure to discard redundant weights. Our pruning method is simultaneously computationally efficient, provably accurate, and broadly applicable to various network architectures and data distributions. Our empirical comparisons show that our algorithm reliably generates highly compressed networks that incur minimal loss in performance relative to that of the original network. We present experimental results that demonstrate our algorithm's potential to unearth essential network connections that can be trained successfully in isolation, which may be of independent interest." @default.
- W2979789219 created "2019-10-18" @default.
- W2979789219 creator A5006739892 @default.
- W2979789219 creator A5064147145 @default.
- W2979789219 creator A5066830185 @default.
- W2979789219 creator A5078356081 @default.
- W2979789219 creator A5082255004 @default.
- W2979789219 date "2019-10-11" @default.
- W2979789219 modified "2023-09-27" @default.
- W2979789219 title "SiPPing Neural Networks: Sensitivity-informed Provable Pruning of Neural Networks." @default.
- W2979789219 cites W1919191429 @default.
- W2979789219 cites W1981773323 @default.
- W2979789219 cites W1993119087 @default.
- W2979789219 cites W2114766824 @default.
- W2979789219 cites W2117539524 @default.
- W2979789219 cites W2145607950 @default.
- W2979789219 cites W2161591461 @default.
- W2979789219 cites W2167215970 @default.
- W2979789219 cites W2173418170 @default.
- W2979789219 cites W2182629226 @default.
- W2979789219 cites W2194775991 @default.
- W2979789219 cites W2279098554 @default.
- W2979789219 cites W2401231614 @default.
- W2979789219 cites W2559873021 @default.
- W2979789219 cites W2707890836 @default.
- W2979789219 cites W2752037867 @default.
- W2979789219 cites W2754084392 @default.
- W2979789219 cites W2786054724 @default.
- W2979789219 cites W2912260645 @default.
- W2979789219 cites W2915589364 @default.
- W2979789219 cites W2950967261 @default.
- W2979789219 cites W2962835968 @default.
- W2979789219 cites W2962950822 @default.
- W2979789219 cites W2962963202 @default.
- W2979789219 cites W2962988160 @default.
- W2979789219 cites W2963000224 @default.
- W2979789219 cites W2963145730 @default.
- W2979789219 cites W2963225922 @default.
- W2979789219 cites W2963236897 @default.
- W2979789219 cites W2963239103 @default.
- W2979789219 cites W2963247446 @default.
- W2979789219 cites W2963285844 @default.
- W2979789219 cites W2963331432 @default.
- W2979789219 cites W2963433148 @default.
- W2979789219 cites W2963446712 @default.
- W2979789219 cites W2963452728 @default.
- W2979789219 cites W2963479620 @default.
- W2979789219 cites W2964089577 @default.
- W2979789219 cites W2964152344 @default.
- W2979789219 cites W2964299589 @default.
- W2979789219 cites W2965862774 @default.
- W2979789219 cites W2970330753 @default.
- W2979789219 cites W2989808579 @default.
- W2979789219 cites W2995428361 @default.
- W2979789219 cites W2995463996 @default.
- W2979789219 cites W2996603747 @default.
- W2979789219 cites W3000181403 @default.
- W2979789219 cites W3038041907 @default.
- W2979789219 cites W566555209 @default.
- W2979789219 hasPublicationYear "2019" @default.
- W2979789219 type Work @default.
- W2979789219 sameAs 2979789219 @default.
- W2979789219 citedByCount "11" @default.
- W2979789219 countsByYear W29797892192019 @default.
- W2979789219 countsByYear W29797892192020 @default.
- W2979789219 countsByYear W29797892192021 @default.
- W2979789219 crossrefType "posted-content" @default.
- W2979789219 hasAuthorship W2979789219A5006739892 @default.
- W2979789219 hasAuthorship W2979789219A5064147145 @default.
- W2979789219 hasAuthorship W2979789219A5066830185 @default.
- W2979789219 hasAuthorship W2979789219A5078356081 @default.
- W2979789219 hasAuthorship W2979789219A5082255004 @default.
- W2979789219 hasConcept C106131492 @default.
- W2979789219 hasConcept C108010975 @default.
- W2979789219 hasConcept C11413529 @default.
- W2979789219 hasConcept C119857082 @default.
- W2979789219 hasConcept C140779682 @default.
- W2979789219 hasConcept C154945302 @default.
- W2979789219 hasConcept C199360897 @default.
- W2979789219 hasConcept C2780801425 @default.
- W2979789219 hasConcept C31972630 @default.
- W2979789219 hasConcept C41008148 @default.
- W2979789219 hasConcept C50644808 @default.
- W2979789219 hasConcept C6557445 @default.
- W2979789219 hasConcept C86803240 @default.
- W2979789219 hasConceptScore W2979789219C106131492 @default.
- W2979789219 hasConceptScore W2979789219C108010975 @default.
- W2979789219 hasConceptScore W2979789219C11413529 @default.
- W2979789219 hasConceptScore W2979789219C119857082 @default.
- W2979789219 hasConceptScore W2979789219C140779682 @default.
- W2979789219 hasConceptScore W2979789219C154945302 @default.
- W2979789219 hasConceptScore W2979789219C199360897 @default.
- W2979789219 hasConceptScore W2979789219C2780801425 @default.
- W2979789219 hasConceptScore W2979789219C31972630 @default.
- W2979789219 hasConceptScore W2979789219C41008148 @default.
- W2979789219 hasConceptScore W2979789219C50644808 @default.
- W2979789219 hasConceptScore W2979789219C6557445 @default.
- W2979789219 hasConceptScore W2979789219C86803240 @default.
- W2979789219 hasLocation W29797892191 @default.
- W2979789219 hasOpenAccess W2979789219 @default.