Federated learning aggregation can be improved by weighting clients with label diversity and model divergence, but tuning the priority order on test accuracy overstates the benefit.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2019 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Towards Effective Device-Aware Federated Learning
Federated learning aggregation can be improved by weighting clients with label diversity and model divergence, but tuning the priority order on test accuracy overstates the benefit.