TY - GEN
T1 - REFL
T2 - 18th European Conference on Computer Systems, EuroSys 2023
AU - Abdelmoniem, Ahmed M.
AU - Sahu, Atal Narayan
AU - Canini, Marco
AU - Fahmy, Suhaib A.
N1 - Publisher Copyright:
© 2023 Copyright held by the owner/author(s). Publication rights licensed to ACM.
PY - 2023/5/8
Y1 - 2023/5/8
N2 - Federated Learning (FL) enables distributed training by learners using local data, thereby enhancing privacy and reducing communication. However, it presents numerous challenges relating to the heterogeneity of the data distribution, device capabilities, and participant availability as deployments scale, which can impact both model convergence and bias. Existing FL schemes use random participant selection to improve the fairness of the selection process; however, this can result in inefficient use of resources and lower quality training. In this work, we systematically address the question of resource efficiency in FL, showing the benefits of intelligent participant selection, and incorporation of updates from straggling participants. We demonstrate how these factors enable resource efficiency while also improving trained model quality.
AB - Federated Learning (FL) enables distributed training by learners using local data, thereby enhancing privacy and reducing communication. However, it presents numerous challenges relating to the heterogeneity of the data distribution, device capabilities, and participant availability as deployments scale, which can impact both model convergence and bias. Existing FL schemes use random participant selection to improve the fairness of the selection process; however, this can result in inefficient use of resources and lower quality training. In this work, we systematically address the question of resource efficiency in FL, showing the benefits of intelligent participant selection, and incorporation of updates from straggling participants. We demonstrate how these factors enable resource efficiency while also improving trained model quality.
UR - http://www.scopus.com/inward/record.url?scp=85152521762&partnerID=8YFLogxK
U2 - 10.1145/3552326.3567485
DO - 10.1145/3552326.3567485
M3 - Conference contribution
AN - SCOPUS:85152521762
T3 - Proceedings of the 18th European Conference on Computer Systems, EuroSys 2023
SP - 215
EP - 232
BT - Proceedings of the 18th European Conference on Computer Systems, EuroSys 2023
PB - Association for Computing Machinery, Inc
Y2 - 8 May 2023 through 12 May 2023
ER -