ProxyRL-FRS blends a device-side proxy model that predicts a client's training loss with a staleness-aware reinforcement learning reward to select clients in federated recommendation, reporting accuracy gains on MovieLens-1M, Fashion, and Video Games.
Federated recommenders: methods, challenges and future,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.IR 1years
2025 1verdicts
CONDITIONAL 1roles
background 1polarities
unclear 1representative citing papers
citing papers explorer
-
Proxy Model-Guided Reinforcement Learning for Client Selection in Federated Recommendation
ProxyRL-FRS blends a device-side proxy model that predicts a client's training loss with a staleness-aware reinforcement learning reward to select clients in federated recommendation, reporting accuracy gains on MovieLens-1M, Fashion, and Video Games.