FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W2897110360> ?p ?o ?g. }

• W2897110360 abstract "Pruning is a popular method for compressing a neural network: given a large trained network, one alternates between removing connections and fine-tuning; reducing the overall width of the network. However, the efficacy of network pruning has largely evaded scrutiny. In this paper, we examine ResNets and DenseNets obtained through pruning-and-tuning and make two interesting observations: (i) reduced networks---smaller versions of the original network trained from scratch---consistently outperform pruned networks; (ii) if you take the architecture of a pruned network and then train it from scratch it is significantly more competitive. Furthermore, these architectures are easy to approximate: we can prune once and obtain a whole family of new, scalable network architectures that can simply be trained from scratch. Finally, we compare the inference speed of reduced and pruned networks on hardware, and show that reduced networks are significantly faster." @default.
• W2897110360 created "2018-10-26" @default.
• W2897110360 creator A5003510948 @default.
• W2897110360 creator A5007901825 @default.
• W2897110360 creator A5027001025 @default.
• W2897110360 creator A5049917315 @default.
• W2897110360 date "2018-10-10" @default.
• W2897110360 modified "2023-09-27" @default.
• W2897110360 title "Pruning neural networks: is it time to nip it in the bud?" @default.
• W2897110360 cites W1821462560 @default.
• W2897110360 cites W2114766824 @default.
• W2897110360 cites W2117539524 @default.
• W2897110360 cites W2120972216 @default.
• W2897110360 cites W2134797427 @default.
• W2897110360 cites W2194775991 @default.
• W2897110360 cites W2401231614 @default.
• W2897110360 cites W2460144244 @default.
• W2897110360 cites W2513419314 @default.
• W2897110360 cites W2541674938 @default.
• W2897110360 cites W2604319603 @default.
• W2897110360 cites W2617242334 @default.
• W2897110360 cites W2619890685 @default.
• W2897110360 cites W2737121650 @default.
• W2897110360 cites W2765720631 @default.
• W2897110360 cites W2768607686 @default.
• W2897110360 cites W2783873922 @default.
• W2897110360 cites W2786054724 @default.
• W2897110360 cites W2810075754 @default.
• W2897110360 cites W2895171208 @default.
• W2897110360 cites W2901394229 @default.
• W2897110360 cites W2915589364 @default.
• W2897110360 cites W2919115771 @default.
• W2897110360 cites W2947776571 @default.
• W2897110360 cites W2949117887 @default.
• W2897110360 cites W2949298282 @default.
• W2897110360 cites W2949829435 @default.
• W2897110360 cites W2949892913 @default.
• W2897110360 cites W2949941638 @default.
• W2897110360 cites W2951569836 @default.
• W2897110360 cites W2953212265 @default.
• W2897110360 cites W2962819303 @default.
• W2897110360 cites W2962861284 @default.
• W2897110360 cites W2962935923 @default.
• W2897110360 cites W2963247446 @default.
• W2897110360 cites W2963382180 @default.
• W2897110360 cites W2963446712 @default.
• W2897110360 cites W2963674932 @default.
• W2897110360 cites W2963813662 @default.
• W2897110360 cites W2963840672 @default.
• W2897110360 cites W2964081807 @default.
• W2897110360 cites W2964299589 @default.
• W2897110360 cites W3118608800 @default.
• W2897110360 hasPublicationYear "2018" @default.
• W2897110360 type Work @default.
• W2897110360 sameAs 2897110360 @default.
• W2897110360 citedByCount "15" @default.
• W2897110360 countsByYear W28971103602018 @default.
• W2897110360 countsByYear W28971103602019 @default.
• W2897110360 countsByYear W28971103602020 @default.
• W2897110360 countsByYear W28971103602021 @default.
• W2897110360 crossrefType "posted-content" @default.
• W2897110360 hasAuthorship W2897110360A5003510948 @default.
• W2897110360 hasAuthorship W2897110360A5007901825 @default.
• W2897110360 hasAuthorship W2897110360A5027001025 @default.
• W2897110360 hasAuthorship W2897110360A5049917315 @default.
• W2897110360 hasConcept C108010975 @default.
• W2897110360 hasConcept C111919701 @default.
• W2897110360 hasConcept C154945302 @default.
• W2897110360 hasConcept C193415008 @default.
• W2897110360 hasConcept C2776214188 @default.
• W2897110360 hasConcept C2781235140 @default.
• W2897110360 hasConcept C31258907 @default.
• W2897110360 hasConcept C41008148 @default.
• W2897110360 hasConcept C48044578 @default.
• W2897110360 hasConcept C50644808 @default.
• W2897110360 hasConcept C6557445 @default.
• W2897110360 hasConcept C77088390 @default.
• W2897110360 hasConcept C86803240 @default.
• W2897110360 hasConceptScore W2897110360C108010975 @default.
• W2897110360 hasConceptScore W2897110360C111919701 @default.
• W2897110360 hasConceptScore W2897110360C154945302 @default.
• W2897110360 hasConceptScore W2897110360C193415008 @default.
• W2897110360 hasConceptScore W2897110360C2776214188 @default.
• W2897110360 hasConceptScore W2897110360C2781235140 @default.
• W2897110360 hasConceptScore W2897110360C31258907 @default.
• W2897110360 hasConceptScore W2897110360C41008148 @default.
• W2897110360 hasConceptScore W2897110360C48044578 @default.
• W2897110360 hasConceptScore W2897110360C50644808 @default.
• W2897110360 hasConceptScore W2897110360C6557445 @default.
• W2897110360 hasConceptScore W2897110360C77088390 @default.
• W2897110360 hasConceptScore W2897110360C86803240 @default.
• W2897110360 hasLocation W28971103601 @default.
• W2897110360 hasOpenAccess W2897110360 @default.
• W2897110360 hasPrimaryLocation W28971103601 @default.
• W2897110360 hasRelatedWork W2068837659 @default.
• W2897110360 hasRelatedWork W2114766824 @default.
• W2897110360 hasRelatedWork W2125389748 @default.
• W2897110360 hasRelatedWork W2194775991 @default.
• W2897110360 hasRelatedWork W2612445135 @default.
• W2897110360 hasRelatedWork W2752037867 @default.

• next