FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W2897824775> ?p ?o ?g. }

• W2897824775 abstract "Deep neural networks (DNNs) although achieving human-level performance in many domains, have very large model size that hinders their broader applications on edge computing devices. Extensive research work have been conducted on DNN model compression or pruning. However, most of the previous work took heuristic approaches. This work proposes a progressive weight pruning approach based on ADMM (Alternating Direction Method of Multipliers), a powerful technique to deal with non-convex optimization problems with potentially combinatorial constraints. Motivated by dynamic programming, the proposed method reaches extremely high pruning rate by using partial prunings with moderate pruning rates. Therefore, it resolves the accuracy degradation and long convergence time problems when pursuing extremely high pruning ratios. It achieves up to 34 times pruning rate for ImageNet dataset and 167 times pruning rate for MNIST dataset, significantly higher than those reached by the literature work. Under the same number of epochs, the proposed method also achieves faster convergence and higher compression rates. The codes and pruned DNN models are released in the link bit.ly/2zxdlss" @default.
• W2897824775 created "2018-10-26" @default.
• W2897824775 creator A5002976916 @default.
• W2897824775 creator A5010753766 @default.
• W2897824775 creator A5011173717 @default.
• W2897824775 creator A5016139042 @default.
• W2897824775 creator A5025596795 @default.
• W2897824775 creator A5043582832 @default.
• W2897824775 creator A5057639149 @default.
• W2897824775 creator A5060314107 @default.
• W2897824775 creator A5060388088 @default.
• W2897824775 creator A5072440587 @default.
• W2897824775 creator A5072864969 @default.
• W2897824775 creator A5076768386 @default.
• W2897824775 creator A5077231313 @default.
• W2897824775 date "2018-10-16" @default.
• W2897824775 modified "2023-09-25" @default.
• W2897824775 title "Progressive Weight Pruning of Deep Neural Networks using ADMM" @default.
• W2897824775 cites W1902934009 @default.
• W2897824775 cites W2108598243 @default.
• W2897824775 cites W2112796928 @default.
• W2897824775 cites W2147768505 @default.
• W2897824775 cites W2160815625 @default.
• W2897824775 cites W2164278908 @default.
• W2897824775 cites W2194775991 @default.
• W2897824775 cites W2233116163 @default.
• W2897824775 cites W2513419314 @default.
• W2897824775 cites W2560017826 @default.
• W2897824775 cites W2593245696 @default.
• W2897824775 cites W2619096655 @default.
• W2897824775 cites W2719597717 @default.
• W2897824775 cites W2737121650 @default.
• W2897824775 cites W2739789140 @default.
• W2897824775 cites W2748818695 @default.
• W2897824775 cites W2754084392 @default.
• W2897824775 cites W2798170643 @default.
• W2897824775 cites W2884180697 @default.
• W2897824775 cites W2952055246 @default.
• W2897824775 cites W2952746978 @default.
• W2897824775 cites W2954582488 @default.
• W2897824775 cites W2962835968 @default.
• W2897824775 cites W2963319203 @default.
• W2897824775 cites W2963452728 @default.
• W2897824775 cites W2963674932 @default.
• W2897824775 cites W2964299589 @default.
• W2897824775 doi "https://doi.org/10.48550/arxiv.1810.07378" @default.
• W2897824775 hasPublicationYear "2018" @default.
• W2897824775 type Work @default.
• W2897824775 sameAs 2897824775 @default.
• W2897824775 citedByCount "18" @default.
• W2897824775 countsByYear W28978247752018 @default.
• W2897824775 countsByYear W28978247752019 @default.
• W2897824775 countsByYear W28978247752020 @default.
• W2897824775 countsByYear W28978247752021 @default.
• W2897824775 crossrefType "posted-content" @default.
• W2897824775 hasAuthorship W2897824775A5002976916 @default.
• W2897824775 hasAuthorship W2897824775A5010753766 @default.
• W2897824775 hasAuthorship W2897824775A5011173717 @default.
• W2897824775 hasAuthorship W2897824775A5016139042 @default.
• W2897824775 hasAuthorship W2897824775A5025596795 @default.
• W2897824775 hasAuthorship W2897824775A5043582832 @default.
• W2897824775 hasAuthorship W2897824775A5057639149 @default.
• W2897824775 hasAuthorship W2897824775A5060314107 @default.
• W2897824775 hasAuthorship W2897824775A5060388088 @default.
• W2897824775 hasAuthorship W2897824775A5072440587 @default.
• W2897824775 hasAuthorship W2897824775A5072864969 @default.
• W2897824775 hasAuthorship W2897824775A5076768386 @default.
• W2897824775 hasAuthorship W2897824775A5077231313 @default.
• W2897824775 hasBestOaLocation W28978247751 @default.
• W2897824775 hasConcept C108010975 @default.
• W2897824775 hasConcept C11413529 @default.
• W2897824775 hasConcept C119857082 @default.
• W2897824775 hasConcept C126255220 @default.
• W2897824775 hasConcept C127162648 @default.
• W2897824775 hasConcept C154945302 @default.
• W2897824775 hasConcept C162307627 @default.
• W2897824775 hasConcept C162324750 @default.
• W2897824775 hasConcept C173801870 @default.
• W2897824775 hasConcept C190502265 @default.
• W2897824775 hasConcept C2777303404 @default.
• W2897824775 hasConcept C2984842247 @default.
• W2897824775 hasConcept C33923547 @default.
• W2897824775 hasConcept C41008148 @default.
• W2897824775 hasConcept C50522688 @default.
• W2897824775 hasConcept C50644808 @default.
• W2897824775 hasConcept C57869625 @default.
• W2897824775 hasConcept C6557445 @default.
• W2897824775 hasConcept C76155785 @default.
• W2897824775 hasConcept C86803240 @default.
• W2897824775 hasConceptScore W2897824775C108010975 @default.
• W2897824775 hasConceptScore W2897824775C11413529 @default.
• W2897824775 hasConceptScore W2897824775C119857082 @default.
• W2897824775 hasConceptScore W2897824775C126255220 @default.
• W2897824775 hasConceptScore W2897824775C127162648 @default.
• W2897824775 hasConceptScore W2897824775C154945302 @default.
• W2897824775 hasConceptScore W2897824775C162307627 @default.
• W2897824775 hasConceptScore W2897824775C162324750 @default.
• W2897824775 hasConceptScore W2897824775C173801870 @default.
• W2897824775 hasConceptScore W2897824775C190502265 @default.
• W2897824775 hasConceptScore W2897824775C2777303404 @default.

• next