Weighted robust aggregation plus a double-momentum variance-reduction method achieves a dimension-free, O(1/sqrt(T)) excess-loss rate for Byzantine-robust asynchronous convex optimization.
Asynchronous distributed learning: Adapting to gradient delays without prior knowledge
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.LG 1years
2025 1verdicts
CONDITIONAL 1roles
background 1polarities
background 1representative citing papers
citing papers explorer
-
Weight for Robustness: A Comprehensive Approach towards Optimal Fault-Tolerant Asynchronous ML
Weighted robust aggregation plus a double-momentum variance-reduction method achieves a dimension-free, O(1/sqrt(T)) excess-loss rate for Byzantine-robust asynchronous convex optimization.