Semi-decentralized federated learning with client pairing for efficient mutual knowledge transfer

Abstract In Decentralized Federated Learning (DFL), Deep Mutual Learning (DML) improves global accuracy under non-independent and identically distributed (non-IID) data by enabling knowledge exchange over clients, but introduces extra training overhead and delays convergence. To solve this issue, we...

Descrizione completa

Dettagli Bibliografici

Autori principali:	Dain Yang, Joohyung Lee, Seong Gon Choi
Natura:	Artigo
Lingua:	Inglês
Pubblicazione:	Nature Portfolio 2025-11-01
Serie:	Scientific Reports
Soggetti:	Data heterogeneity Decentralized federated learning Knowledge transfer
Accesso online:	https://doi.org/10.1038/s41598-025-29491-4
Tags:	Aggiungi Tag Nessun Tag, puoi essere il primo ad aggiungerne!!