Matches in SemOpenAlex for { <https://semopenalex.org/work/W4286571684> ?p ?o ?g. }
Showing items 1 to 95 of
95
with 100 items per page.
- W4286571684 abstract "In decentralized machine learning over a network of workers, each worker updates its local model as a weighted average of its local model and all models received from its neighbors. Efficient consensus weight matrix design and communication resource allocation can increase the training convergence rate and reduce the wall-clock training time. In this paper, we jointly consider these two factors and propose a novel algorithm termed Communication-Efficient Network Topology (CENT), which reduces the latency in each training iteration by removing unnecessary communication links. CENT preserves the training convergence rate while enforcing communication graph sparsity and avoiding selecting poor communication links. Numerical study with real-world machine learning data demonstrates the efficacy of the proposed solution and its performance advantage over state-of-the-art algorithms." @default.
- W4286571684 created "2022-07-22" @default.
- W4286571684 creator A5009603325 @default.
- W4286571684 creator A5040221369 @default.
- W4286571684 creator A5041434153 @default.
- W4286571684 creator A5086734771 @default.
- W4286571684 creator A5087351444 @default.
- W4286571684 date "2022-06-13" @default.
- W4286571684 modified "2023-09-23" @default.
- W4286571684 title "Joint Consensus Matrix Design and Resource Allocation for Decentralized Learning" @default.
- W4286571684 cites W132903907 @default.
- W4286571684 cites W2063403497 @default.
- W4286571684 cites W2074032690 @default.
- W4286571684 cites W2074796812 @default.
- W4286571684 cites W2095057433 @default.
- W4286571684 cites W2101517602 @default.
- W4286571684 cites W2112796928 @default.
- W4286571684 cites W2134711723 @default.
- W4286571684 cites W2138547065 @default.
- W4286571684 cites W2151685478 @default.
- W4286571684 cites W2152244994 @default.
- W4286571684 cites W2167362932 @default.
- W4286571684 cites W2344056755 @default.
- W4286571684 cites W2400861403 @default.
- W4286571684 cites W2418562807 @default.
- W4286571684 cites W2571422703 @default.
- W4286571684 cites W2789364541 @default.
- W4286571684 cites W2920179928 @default.
- W4286571684 cites W2963786636 @default.
- W4286571684 cites W3012125688 @default.
- W4286571684 cites W3016159127 @default.
- W4286571684 cites W3022983677 @default.
- W4286571684 cites W3089411138 @default.
- W4286571684 cites W3105329945 @default.
- W4286571684 cites W3155415667 @default.
- W4286571684 doi "https://doi.org/10.23919/ifipnetworking55013.2022.9829798" @default.
- W4286571684 hasPublicationYear "2022" @default.
- W4286571684 type Work @default.
- W4286571684 citedByCount "0" @default.
- W4286571684 crossrefType "proceedings-article" @default.
- W4286571684 hasAuthorship W4286571684A5009603325 @default.
- W4286571684 hasAuthorship W4286571684A5040221369 @default.
- W4286571684 hasAuthorship W4286571684A5041434153 @default.
- W4286571684 hasAuthorship W4286571684A5086734771 @default.
- W4286571684 hasAuthorship W4286571684A5087351444 @default.
- W4286571684 hasConcept C119857082 @default.
- W4286571684 hasConcept C120314980 @default.
- W4286571684 hasConcept C126255220 @default.
- W4286571684 hasConcept C127162648 @default.
- W4286571684 hasConcept C132525143 @default.
- W4286571684 hasConcept C154945302 @default.
- W4286571684 hasConcept C162324750 @default.
- W4286571684 hasConcept C199845137 @default.
- W4286571684 hasConcept C2777303404 @default.
- W4286571684 hasConcept C31258907 @default.
- W4286571684 hasConcept C33923547 @default.
- W4286571684 hasConcept C41008148 @default.
- W4286571684 hasConcept C50522688 @default.
- W4286571684 hasConcept C57869625 @default.
- W4286571684 hasConcept C76155785 @default.
- W4286571684 hasConcept C80444323 @default.
- W4286571684 hasConcept C82876162 @default.
- W4286571684 hasConceptScore W4286571684C119857082 @default.
- W4286571684 hasConceptScore W4286571684C120314980 @default.
- W4286571684 hasConceptScore W4286571684C126255220 @default.
- W4286571684 hasConceptScore W4286571684C127162648 @default.
- W4286571684 hasConceptScore W4286571684C132525143 @default.
- W4286571684 hasConceptScore W4286571684C154945302 @default.
- W4286571684 hasConceptScore W4286571684C162324750 @default.
- W4286571684 hasConceptScore W4286571684C199845137 @default.
- W4286571684 hasConceptScore W4286571684C2777303404 @default.
- W4286571684 hasConceptScore W4286571684C31258907 @default.
- W4286571684 hasConceptScore W4286571684C33923547 @default.
- W4286571684 hasConceptScore W4286571684C41008148 @default.
- W4286571684 hasConceptScore W4286571684C50522688 @default.
- W4286571684 hasConceptScore W4286571684C57869625 @default.
- W4286571684 hasConceptScore W4286571684C76155785 @default.
- W4286571684 hasConceptScore W4286571684C80444323 @default.
- W4286571684 hasConceptScore W4286571684C82876162 @default.
- W4286571684 hasLocation W42865716841 @default.
- W4286571684 hasOpenAccess W4286571684 @default.
- W4286571684 hasPrimaryLocation W42865716841 @default.
- W4286571684 hasRelatedWork W14296039 @default.
- W4286571684 hasRelatedWork W1566886392 @default.
- W4286571684 hasRelatedWork W2071793248 @default.
- W4286571684 hasRelatedWork W2079802744 @default.
- W4286571684 hasRelatedWork W2107367999 @default.
- W4286571684 hasRelatedWork W2152241334 @default.
- W4286571684 hasRelatedWork W2298102683 @default.
- W4286571684 hasRelatedWork W2385763152 @default.
- W4286571684 hasRelatedWork W2919369587 @default.
- W4286571684 hasRelatedWork W2968850319 @default.
- W4286571684 isParatext "false" @default.
- W4286571684 isRetracted "false" @default.
- W4286571684 workType "article" @default.