FedAPA adaptively learns personalized aggregation weights on the server, using the change in each client's model weights as a gradient signal, and reports top accuracy across 10 of 12 non-IID FL benchmarks.
Fedphp: Federated personalization with inherited private models
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
REJECT 1representative citing papers
citing papers explorer
-
FedAPA: Server-side Gradient-Based Adaptive Personalized Aggregation for Federated Learning on Heterogeneous Data
FedAPA adaptively learns personalized aggregation weights on the server, using the change in each client's model weights as a gradient signal, and reports top accuracy across 10 of 12 non-IID FL benchmarks.