Matches in SemOpenAlex for { <https://semopenalex.org/work/W4385681762> ?p ?o ?g. }
Showing items 1 to 75 of
75
with 100 items per page.
- W4385681762 abstract "Federated learning (FL) is a privacy-preserving collaborative learning framework, and differential privacy can be applied to further enhance its privacy protection. Existing FL systems typically adopt Federated Average (FedAvg) as the training algorithm and implement differential privacy with a Gaussian mechanism. However, the inherent privacy-utility trade-off in these systems severely degrades the training performance if a tight privacy budget is enforced. Besides, the Gaussian mechanism requires model weights to be of high-precision. To improve communication efficiency and achieve a better privacy-utility trade-off, we propose a communication-efficient FL training algorithm with differential privacy guarantee. Specifically, we propose to adopt binary neural networks (BNNs) and introduce discrete noise in the FL setting. Binary model parameters are uploaded for higher communication efficiency and discrete noise is added to achieve the client-level differential privacy protection. The achieved performance guarantee is rigorously proved, and it is shown to depend on the level of discrete noise. Experimental results based on MNIST and Fashion-MNIST datasets will demonstrate that the proposed training algorithm achieves client-level privacy protection with performance gain while enjoying the benefits of low communication overhead from binary model updates." @default.
- W4385681762 created "2023-08-09" @default.
- W4385681762 creator A5009944050 @default.
- W4385681762 creator A5025467937 @default.
- W4385681762 creator A5071672663 @default.
- W4385681762 creator A5079052203 @default.
- W4385681762 date "2023-08-07" @default.
- W4385681762 modified "2023-10-16" @default.
- W4385681762 title "Binary Federated Learning with Client-Level Differential Privacy" @default.
- W4385681762 doi "https://doi.org/10.48550/arxiv.2308.03320" @default.
- W4385681762 hasPublicationYear "2023" @default.
- W4385681762 type Work @default.
- W4385681762 citedByCount "0" @default.
- W4385681762 crossrefType "posted-content" @default.
- W4385681762 hasAuthorship W4385681762A5009944050 @default.
- W4385681762 hasAuthorship W4385681762A5025467937 @default.
- W4385681762 hasAuthorship W4385681762A5071672663 @default.
- W4385681762 hasAuthorship W4385681762A5079052203 @default.
- W4385681762 hasBestOaLocation W43856817621 @default.
- W4385681762 hasConcept C111919701 @default.
- W4385681762 hasConcept C11413529 @default.
- W4385681762 hasConcept C115961682 @default.
- W4385681762 hasConcept C119857082 @default.
- W4385681762 hasConcept C124101348 @default.
- W4385681762 hasConcept C127413603 @default.
- W4385681762 hasConcept C146978453 @default.
- W4385681762 hasConcept C154945302 @default.
- W4385681762 hasConcept C190502265 @default.
- W4385681762 hasConcept C23130292 @default.
- W4385681762 hasConcept C2779960059 @default.
- W4385681762 hasConcept C33923547 @default.
- W4385681762 hasConcept C38652104 @default.
- W4385681762 hasConcept C41008148 @default.
- W4385681762 hasConcept C48372109 @default.
- W4385681762 hasConcept C50644808 @default.
- W4385681762 hasConcept C71901391 @default.
- W4385681762 hasConcept C93226319 @default.
- W4385681762 hasConcept C94375191 @default.
- W4385681762 hasConcept C99498987 @default.
- W4385681762 hasConceptScore W4385681762C111919701 @default.
- W4385681762 hasConceptScore W4385681762C11413529 @default.
- W4385681762 hasConceptScore W4385681762C115961682 @default.
- W4385681762 hasConceptScore W4385681762C119857082 @default.
- W4385681762 hasConceptScore W4385681762C124101348 @default.
- W4385681762 hasConceptScore W4385681762C127413603 @default.
- W4385681762 hasConceptScore W4385681762C146978453 @default.
- W4385681762 hasConceptScore W4385681762C154945302 @default.
- W4385681762 hasConceptScore W4385681762C190502265 @default.
- W4385681762 hasConceptScore W4385681762C23130292 @default.
- W4385681762 hasConceptScore W4385681762C2779960059 @default.
- W4385681762 hasConceptScore W4385681762C33923547 @default.
- W4385681762 hasConceptScore W4385681762C38652104 @default.
- W4385681762 hasConceptScore W4385681762C41008148 @default.
- W4385681762 hasConceptScore W4385681762C48372109 @default.
- W4385681762 hasConceptScore W4385681762C50644808 @default.
- W4385681762 hasConceptScore W4385681762C71901391 @default.
- W4385681762 hasConceptScore W4385681762C93226319 @default.
- W4385681762 hasConceptScore W4385681762C94375191 @default.
- W4385681762 hasConceptScore W4385681762C99498987 @default.
- W4385681762 hasLocation W43856817621 @default.
- W4385681762 hasOpenAccess W4385681762 @default.
- W4385681762 hasPrimaryLocation W43856817621 @default.
- W4385681762 hasRelatedWork W1983811306 @default.
- W4385681762 hasRelatedWork W2379353248 @default.
- W4385681762 hasRelatedWork W2597787948 @default.
- W4385681762 hasRelatedWork W3046565236 @default.
- W4385681762 hasRelatedWork W3101456894 @default.
- W4385681762 hasRelatedWork W4226170255 @default.
- W4385681762 hasRelatedWork W4313446918 @default.
- W4385681762 hasRelatedWork W4367334914 @default.
- W4385681762 hasRelatedWork W4379517515 @default.
- W4385681762 hasRelatedWork W4385489274 @default.
- W4385681762 isParatext "false" @default.
- W4385681762 isRetracted "false" @default.
- W4385681762 workType "article" @default.