OLALa adapts lattice quantizers per client and per round in federated learning, preserving the O(1/T) convergence rate and improving accuracy over fixed lattice baselines.
Federated learning: A signal processing perspective,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
eess.SP 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
OLALa: Online Learned Adaptive Lattice Codes for Heterogeneous Federated Learning
OLALa adapts lattice quantizers per client and per round in federated learning, preserving the O(1/T) convergence rate and improving accuracy over fixed lattice baselines.