Matches in SemOpenAlex for { <https://semopenalex.org/work/W2918777711> ?p ?o ?g. }
- W2918777711 abstract "Federated Learning allows multiple parties to jointly train a deep learning model on their combined data, without any of the participants having to reveal their local data to a centralized server. This form of privacy-preserving collaborative learning however comes at the cost of a significant communication overhead during training. To address this problem, several compression methods have been proposed in the distributed training literature that can reduce the amount of required communication by up to three orders of magnitude. These existing methods however are only of limited utility in the Federated Learning setting, as they either only compress the upstream communication from the clients to the server (leaving the downstream communication uncompressed) or only perform well under idealized conditions such as iid distribution of the client data, which typically can not be found in Federated Learning. In this work, we propose Sparse Ternary Compression (STC), a new compression framework that is specifically designed to meet the requirements of the Federated Learning environment. Our experiments on four different learning tasks demonstrate that STC distinctively outperforms Federated Averaging in common Federated Learning scenarios where clients either a) hold non-iid data, b) use small batch sizes during training, or where c) the number of clients is large and the participation rate in every communication round is low. We furthermore show that even if the clients hold iid data and use medium sized batches for training, STC still behaves pareto-superior to Federated Averaging in the sense that it achieves fixed target accuracies on our benchmarks within both fewer training iterations and a smaller communication budget." @default.
- W2918777711 created "2019-03-11" @default.
- W2918777711 creator A5009304455 @default.
- W2918777711 creator A5013534474 @default.
- W2918777711 creator A5026451495 @default.
- W2918777711 creator A5072994165 @default.
- W2918777711 date "2019-03-07" @default.
- W2918777711 modified "2023-09-27" @default.
- W2918777711 title "Robust and Communication-Efficient Federated Learning from Non-IID Data" @default.
- W2918777711 cites W1686810756 @default.
- W2918777711 cites W1905882502 @default.
- W2918777711 cites W2016053056 @default.
- W2918777711 cites W2113839990 @default.
- W2918777711 cites W2115613939 @default.
- W2918777711 cites W2130942839 @default.
- W2918777711 cites W2194775991 @default.
- W2918777711 cites W2199499977 @default.
- W2918777711 cites W2283463896 @default.
- W2918777711 cites W2308045930 @default.
- W2918777711 cites W2405578611 @default.
- W2918777711 cites W2473418344 @default.
- W2918777711 cites W2535838896 @default.
- W2918777711 cites W2541884796 @default.
- W2918777711 cites W2750384547 @default.
- W2918777711 cites W2767079719 @default.
- W2918777711 cites W2769644379 @default.
- W2918777711 cites W2773194476 @default.
- W2918777711 cites W2785456003 @default.
- W2918777711 cites W2797583228 @default.
- W2918777711 cites W2803110246 @default.
- W2918777711 cites W2805997383 @default.
- W2918777711 cites W2807006176 @default.
- W2918777711 cites W2810065831 @default.
- W2918777711 cites W2905121006 @default.
- W2918777711 cites W2908007030 @default.
- W2918777711 cites W2913570153 @default.
- W2918777711 cites W2919115771 @default.
- W2918777711 cites W2953012738 @default.
- W2918777711 cites W2963300719 @default.
- W2918777711 cites W2963446712 @default.
- W2918777711 cites W2963664311 @default.
- W2918777711 cites W2963803379 @default.
- W2918777711 cites W2964004663 @default.
- W2918777711 cites W2978015420 @default.
- W2918777711 cites W3101036738 @default.
- W2918777711 hasPublicationYear "2019" @default.
- W2918777711 type Work @default.
- W2918777711 sameAs 2918777711 @default.
- W2918777711 citedByCount "26" @default.
- W2918777711 countsByYear W29187777112019 @default.
- W2918777711 countsByYear W29187777112020 @default.
- W2918777711 countsByYear W29187777112021 @default.
- W2918777711 crossrefType "posted-content" @default.
- W2918777711 hasAuthorship W2918777711A5009304455 @default.
- W2918777711 hasAuthorship W2918777711A5013534474 @default.
- W2918777711 hasAuthorship W2918777711A5026451495 @default.
- W2918777711 hasAuthorship W2918777711A5072994165 @default.
- W2918777711 hasConcept C111919701 @default.
- W2918777711 hasConcept C119857082 @default.
- W2918777711 hasConcept C120314980 @default.
- W2918777711 hasConcept C136764020 @default.
- W2918777711 hasConcept C154945302 @default.
- W2918777711 hasConcept C191172861 @default.
- W2918777711 hasConcept C2779960059 @default.
- W2918777711 hasConcept C2992525071 @default.
- W2918777711 hasConcept C31258907 @default.
- W2918777711 hasConcept C41008148 @default.
- W2918777711 hasConcept C71901391 @default.
- W2918777711 hasConceptScore W2918777711C111919701 @default.
- W2918777711 hasConceptScore W2918777711C119857082 @default.
- W2918777711 hasConceptScore W2918777711C120314980 @default.
- W2918777711 hasConceptScore W2918777711C136764020 @default.
- W2918777711 hasConceptScore W2918777711C154945302 @default.
- W2918777711 hasConceptScore W2918777711C191172861 @default.
- W2918777711 hasConceptScore W2918777711C2779960059 @default.
- W2918777711 hasConceptScore W2918777711C2992525071 @default.
- W2918777711 hasConceptScore W2918777711C31258907 @default.
- W2918777711 hasConceptScore W2918777711C41008148 @default.
- W2918777711 hasConceptScore W2918777711C71901391 @default.
- W2918777711 hasLocation W29187777111 @default.
- W2918777711 hasOpenAccess W2918777711 @default.
- W2918777711 hasPrimaryLocation W29187777111 @default.
- W2918777711 hasRelatedWork W2112796928 @default.
- W2918777711 hasRelatedWork W2116612304 @default.
- W2918777711 hasRelatedWork W2283463896 @default.
- W2918777711 hasRelatedWork W2530417694 @default.
- W2918777711 hasRelatedWork W2535838896 @default.
- W2918777711 hasRelatedWork W2541884796 @default.
- W2918777711 hasRelatedWork W2767079719 @default.
- W2918777711 hasRelatedWork W2807006176 @default.
- W2918777711 hasRelatedWork W2810065831 @default.
- W2918777711 hasRelatedWork W2900120080 @default.
- W2918777711 hasRelatedWork W2902113386 @default.
- W2918777711 hasRelatedWork W2912213068 @default.
- W2918777711 hasRelatedWork W2913570153 @default.
- W2918777711 hasRelatedWork W2914328083 @default.
- W2918777711 hasRelatedWork W2919115771 @default.
- W2918777711 hasRelatedWork W2963300197 @default.
- W2918777711 hasRelatedWork W2963318081 @default.
- W2918777711 hasRelatedWork W3047304572 @default.