Matches in SemOpenAlex for { <https://semopenalex.org/work/W3138285433> ?p ?o ?g. }
Showing items 1 to 64 of
64
with 100 items per page.
- W3138285433 abstract "In federated learning, workers train local models with their private data sets and only upload local gradients to the remote aggregator. Data privacy is well preserved and parallelism is achieved. In large-scale deep learning tasks, however, frequent interactions between workers and the aggregator to transmit parameters can cause tremendous degradation of system performance in terms of communication costs, the needed number of iterations, the latency of each iteration and the accuracy of the trained model because of system “churns” (i.e., devices frequently joining and leaving the network). Existing research leverages different network topologies to improve the performance of federated learning. In this paper, we propose a novel hybrid network topology design that integrates ring (R) and n-ary tree (T) to provide flexible and adaptive convergecast in federated learning. Specifically, multiple participated peers within one-hop are formed as a local ring to adapt to device dynamics (i.e., “churns”) and carry out local cooperation shuffling; an n-ary convergecast tree is formed from local rings to the aggregator to assure the communication efficiency. Theoretical analysis shows the superiority of the proposed hybrid (R+T) convergecast design in terms of system latency as compared to existing topologies. Prototype-based simulation on CloudLab shows that the hybrid (R+T) design is able to reduce the rounds of iterations while achieving the best model accuracy under system “churns” as compared to the state of the art." @default.
- W3138285433 created "2021-03-29" @default.
- W3138285433 creator A5026874272 @default.
- W3138285433 creator A5028358992 @default.
- W3138285433 creator A5063617425 @default.
- W3138285433 date "2021-01-09" @default.
- W3138285433 modified "2023-09-23" @default.
- W3138285433 title "Efficient Parameter Aggregation in Federated Learning with Hybrid Convergecast" @default.
- W3138285433 cites W2053637704 @default.
- W3138285433 cites W2114900869 @default.
- W3138285433 cites W3105122387 @default.
- W3138285433 cites W564832898 @default.
- W3138285433 doi "https://doi.org/10.1109/ccnc49032.2021.9369497" @default.
- W3138285433 hasPublicationYear "2021" @default.
- W3138285433 type Work @default.
- W3138285433 sameAs 3138285433 @default.
- W3138285433 citedByCount "0" @default.
- W3138285433 crossrefType "proceedings-article" @default.
- W3138285433 hasAuthorship W3138285433A5026874272 @default.
- W3138285433 hasAuthorship W3138285433A5028358992 @default.
- W3138285433 hasAuthorship W3138285433A5063617425 @default.
- W3138285433 hasConcept C119857082 @default.
- W3138285433 hasConcept C120314980 @default.
- W3138285433 hasConcept C136764020 @default.
- W3138285433 hasConcept C154945302 @default.
- W3138285433 hasConcept C180505990 @default.
- W3138285433 hasConcept C199845137 @default.
- W3138285433 hasConcept C24590314 @default.
- W3138285433 hasConcept C31258907 @default.
- W3138285433 hasConcept C41008148 @default.
- W3138285433 hasConcept C71901391 @default.
- W3138285433 hasConcept C76155785 @default.
- W3138285433 hasConcept C82578977 @default.
- W3138285433 hasConcept C82876162 @default.
- W3138285433 hasConceptScore W3138285433C119857082 @default.
- W3138285433 hasConceptScore W3138285433C120314980 @default.
- W3138285433 hasConceptScore W3138285433C136764020 @default.
- W3138285433 hasConceptScore W3138285433C154945302 @default.
- W3138285433 hasConceptScore W3138285433C180505990 @default.
- W3138285433 hasConceptScore W3138285433C199845137 @default.
- W3138285433 hasConceptScore W3138285433C24590314 @default.
- W3138285433 hasConceptScore W3138285433C31258907 @default.
- W3138285433 hasConceptScore W3138285433C41008148 @default.
- W3138285433 hasConceptScore W3138285433C71901391 @default.
- W3138285433 hasConceptScore W3138285433C76155785 @default.
- W3138285433 hasConceptScore W3138285433C82578977 @default.
- W3138285433 hasConceptScore W3138285433C82876162 @default.
- W3138285433 hasLocation W31382854331 @default.
- W3138285433 hasOpenAccess W3138285433 @default.
- W3138285433 hasPrimaryLocation W31382854331 @default.
- W3138285433 hasRelatedWork W1543783347 @default.
- W3138285433 hasRelatedWork W2292891061 @default.
- W3138285433 hasRelatedWork W2298102683 @default.
- W3138285433 hasRelatedWork W2385763152 @default.
- W3138285433 hasRelatedWork W2624382891 @default.
- W3138285433 hasRelatedWork W2752879513 @default.
- W3138285433 hasRelatedWork W2942611981 @default.
- W3138285433 hasRelatedWork W3138285433 @default.
- W3138285433 hasRelatedWork W4229763318 @default.
- W3138285433 hasRelatedWork W4242402958 @default.
- W3138285433 isParatext "false" @default.
- W3138285433 isRetracted "false" @default.
- W3138285433 magId "3138285433" @default.
- W3138285433 workType "article" @default.