Matches in SemOpenAlex for { <https://semopenalex.org/work/W4385890223> ?p ?o ?g. }
Showing items 1 to 69 of
69
with 100 items per page.
- W4385890223 abstract "Federated learning (FL) is a promising approach in distributed learning keeping privacy. However, during the training pipeline of FL, slow or incapable clients (i.e., stragglers) slow down the total training time and degrade performance. System heterogeneity, including heterogeneous computing and network bandwidth, has been addressed to mitigate the impact of stragglers. Previous studies tackle the system heterogeneity by splitting a model into submodels, but with less degree-of-freedom in terms of model architecture. We propose nested federated learning (NeFL), a generalized framework that efficiently divides a model into submodels using both depthwise and widthwise scaling. NeFL is implemented by interpreting forward propagation of models as solving ordinary differential equations (ODEs) with adaptive step sizes. To address the inconsistency that arises when training multiple submodels of different architecture, we decouple a few parameters from parameters being trained for each submodel. NeFL enables resource-constrained clients to effectively join the FL pipeline and the model to be trained with a larger amount of data. Through a series of experiments, we demonstrate that NeFL leads to significant performance gains, especially for the worst-case submodel. Furthermore, we demonstrate NeFL aligns with recent studies in FL, regarding pre-trained models of FL and the statistical heterogeneity." @default.
- W4385890223 created "2023-08-17" @default.
- W4385890223 creator A5035397475 @default.
- W4385890223 creator A5045800916 @default.
- W4385890223 creator A5062994489 @default.
- W4385890223 creator A5077914937 @default.
- W4385890223 creator A5080510562 @default.
- W4385890223 date "2023-08-15" @default.
- W4385890223 modified "2023-10-13" @default.
- W4385890223 title "NeFL: Nested Federated Learning for Heterogeneous Clients" @default.
- W4385890223 doi "https://doi.org/10.48550/arxiv.2308.07761" @default.
- W4385890223 hasPublicationYear "2023" @default.
- W4385890223 type Work @default.
- W4385890223 citedByCount "0" @default.
- W4385890223 crossrefType "posted-content" @default.
- W4385890223 hasAuthorship W4385890223A5035397475 @default.
- W4385890223 hasAuthorship W4385890223A5045800916 @default.
- W4385890223 hasAuthorship W4385890223A5062994489 @default.
- W4385890223 hasAuthorship W4385890223A5077914937 @default.
- W4385890223 hasAuthorship W4385890223A5080510562 @default.
- W4385890223 hasBestOaLocation W43858902231 @default.
- W4385890223 hasConcept C119857082 @default.
- W4385890223 hasConcept C120314980 @default.
- W4385890223 hasConcept C121332964 @default.
- W4385890223 hasConcept C124101348 @default.
- W4385890223 hasConcept C153294291 @default.
- W4385890223 hasConcept C154945302 @default.
- W4385890223 hasConcept C199360897 @default.
- W4385890223 hasConcept C2524010 @default.
- W4385890223 hasConcept C2776257435 @default.
- W4385890223 hasConcept C2777211547 @default.
- W4385890223 hasConcept C31258907 @default.
- W4385890223 hasConcept C33923547 @default.
- W4385890223 hasConcept C41008148 @default.
- W4385890223 hasConcept C43521106 @default.
- W4385890223 hasConcept C80444323 @default.
- W4385890223 hasConcept C99844830 @default.
- W4385890223 hasConceptScore W4385890223C119857082 @default.
- W4385890223 hasConceptScore W4385890223C120314980 @default.
- W4385890223 hasConceptScore W4385890223C121332964 @default.
- W4385890223 hasConceptScore W4385890223C124101348 @default.
- W4385890223 hasConceptScore W4385890223C153294291 @default.
- W4385890223 hasConceptScore W4385890223C154945302 @default.
- W4385890223 hasConceptScore W4385890223C199360897 @default.
- W4385890223 hasConceptScore W4385890223C2524010 @default.
- W4385890223 hasConceptScore W4385890223C2776257435 @default.
- W4385890223 hasConceptScore W4385890223C2777211547 @default.
- W4385890223 hasConceptScore W4385890223C31258907 @default.
- W4385890223 hasConceptScore W4385890223C33923547 @default.
- W4385890223 hasConceptScore W4385890223C41008148 @default.
- W4385890223 hasConceptScore W4385890223C43521106 @default.
- W4385890223 hasConceptScore W4385890223C80444323 @default.
- W4385890223 hasConceptScore W4385890223C99844830 @default.
- W4385890223 hasLocation W43858902231 @default.
- W4385890223 hasOpenAccess W4385890223 @default.
- W4385890223 hasPrimaryLocation W43858902231 @default.
- W4385890223 hasRelatedWork W100620283 @default.
- W4385890223 hasRelatedWork W1530808388 @default.
- W4385890223 hasRelatedWork W230091440 @default.
- W4385890223 hasRelatedWork W2381272478 @default.
- W4385890223 hasRelatedWork W258997015 @default.
- W4385890223 hasRelatedWork W2810751659 @default.
- W4385890223 hasRelatedWork W2997094352 @default.
- W4385890223 hasRelatedWork W3216976533 @default.
- W4385890223 hasRelatedWork W4366179611 @default.
- W4385890223 hasRelatedWork W2233261550 @default.
- W4385890223 isParatext "false" @default.
- W4385890223 isRetracted "false" @default.
- W4385890223 workType "article" @default.