Matches in SemOpenAlex for { <https://semopenalex.org/work/W4387303319> ?p ?o ?g. }
Showing items 1 to 79 of
79
with 100 items per page.
- W4387303319 endingPage "14" @default.
- W4387303319 startingPage "1" @default.
- W4387303319 abstract "Federated learning (FL) is a promising framework for privacy-preserving and distributed training with decentralized clients. However, there exists a large divergence between the collected local updates and the expected global update, which is known as the <italic xmlns:mml=http://www.w3.org/1998/Math/MathML xmlns:xlink=http://www.w3.org/1999/xlink>client drift</i> and mainly caused by heterogeneous data distribution among clients, multiple local training steps, and partial client participation training. Most existing works tackle this challenge based on the empirical risk minimization (ERM) rule, while less attention has been paid to the relationship between the global loss landscape and the generalization ability. In this work, we propose FedGAMMA, a novel FL algorithm with Global sharpness-Aware MiniMizAtion to seek a global flat landscape with high performance. Specifically, in contrast to FedSAM which only seeks the local flatness and still suffers from performance degradation when facing the client-drift issue, we adopt a local varieties control technique to better align each client’s local updates to alleviate the client drift and make each client heading toward the global flatness together. Finally, extensive experiments demonstrate that FedGAMMA can substantially outperform several existing FL baselines on various datasets, and it can well address the <italic xmlns:mml=http://www.w3.org/1998/Math/MathML xmlns:xlink=http://www.w3.org/1999/xlink>client-drift</i> issue and simultaneously seek a smoother and flatter global landscape." @default.
- W4387303319 created "2023-10-04" @default.
- W4387303319 creator A5023307885 @default.
- W4387303319 creator A5034737032 @default.
- W4387303319 creator A5042802004 @default.
- W4387303319 creator A5046305086 @default.
- W4387303319 creator A5049674712 @default.
- W4387303319 creator A5064766287 @default.
- W4387303319 creator A5071510961 @default.
- W4387303319 date "2023-01-01" @default.
- W4387303319 modified "2023-10-14" @default.
- W4387303319 title "FedGAMMA: Federated Learning With Global Sharpness-Aware Minimization" @default.
- W4387303319 doi "https://doi.org/10.1109/tnnls.2023.3304453" @default.
- W4387303319 hasPubMedId "https://pubmed.ncbi.nlm.nih.gov/37788191" @default.
- W4387303319 hasPublicationYear "2023" @default.
- W4387303319 type Work @default.
- W4387303319 citedByCount "0" @default.
- W4387303319 crossrefType "journal-article" @default.
- W4387303319 hasAuthorship W4387303319A5023307885 @default.
- W4387303319 hasAuthorship W4387303319A5034737032 @default.
- W4387303319 hasAuthorship W4387303319A5042802004 @default.
- W4387303319 hasAuthorship W4387303319A5046305086 @default.
- W4387303319 hasAuthorship W4387303319A5049674712 @default.
- W4387303319 hasAuthorship W4387303319A5064766287 @default.
- W4387303319 hasAuthorship W4387303319A5071510961 @default.
- W4387303319 hasConcept C117765406 @default.
- W4387303319 hasConcept C121332964 @default.
- W4387303319 hasConcept C134306372 @default.
- W4387303319 hasConcept C136764020 @default.
- W4387303319 hasConcept C138885662 @default.
- W4387303319 hasConcept C147764199 @default.
- W4387303319 hasConcept C154945302 @default.
- W4387303319 hasConcept C177148314 @default.
- W4387303319 hasConcept C207390915 @default.
- W4387303319 hasConcept C26405456 @default.
- W4387303319 hasConcept C2776502983 @default.
- W4387303319 hasConcept C2778530986 @default.
- W4387303319 hasConcept C33923547 @default.
- W4387303319 hasConcept C41008148 @default.
- W4387303319 hasConcept C41895202 @default.
- W4387303319 hasConcept C50644808 @default.
- W4387303319 hasConcept C62520636 @default.
- W4387303319 hasConceptScore W4387303319C117765406 @default.
- W4387303319 hasConceptScore W4387303319C121332964 @default.
- W4387303319 hasConceptScore W4387303319C134306372 @default.
- W4387303319 hasConceptScore W4387303319C136764020 @default.
- W4387303319 hasConceptScore W4387303319C138885662 @default.
- W4387303319 hasConceptScore W4387303319C147764199 @default.
- W4387303319 hasConceptScore W4387303319C154945302 @default.
- W4387303319 hasConceptScore W4387303319C177148314 @default.
- W4387303319 hasConceptScore W4387303319C207390915 @default.
- W4387303319 hasConceptScore W4387303319C26405456 @default.
- W4387303319 hasConceptScore W4387303319C2776502983 @default.
- W4387303319 hasConceptScore W4387303319C2778530986 @default.
- W4387303319 hasConceptScore W4387303319C33923547 @default.
- W4387303319 hasConceptScore W4387303319C41008148 @default.
- W4387303319 hasConceptScore W4387303319C41895202 @default.
- W4387303319 hasConceptScore W4387303319C50644808 @default.
- W4387303319 hasConceptScore W4387303319C62520636 @default.
- W4387303319 hasFunder F4320321001 @default.
- W4387303319 hasLocation W43873033191 @default.
- W4387303319 hasLocation W43873033192 @default.
- W4387303319 hasOpenAccess W4387303319 @default.
- W4387303319 hasPrimaryLocation W43873033191 @default.
- W4387303319 hasRelatedWork W2060518740 @default.
- W4387303319 hasRelatedWork W2099261031 @default.
- W4387303319 hasRelatedWork W2141212940 @default.
- W4387303319 hasRelatedWork W2381411913 @default.
- W4387303319 hasRelatedWork W3129823125 @default.
- W4387303319 hasRelatedWork W3216201070 @default.
- W4387303319 hasRelatedWork W4221165658 @default.
- W4387303319 hasRelatedWork W4289655432 @default.
- W4387303319 hasRelatedWork W4385007179 @default.
- W4387303319 hasRelatedWork W4386075091 @default.
- W4387303319 isParatext "false" @default.
- W4387303319 isRetracted "false" @default.
- W4387303319 workType "article" @default.