New algorithms achieve regret roughly ||u||G(sqrt(T)+k) under adversarial gradient corruption in unconstrained online convex optimization, with a matching lower bound when G is known.
Gradient descent: Robustness to adversarial corruption
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
unclear 1representative citing papers
citing papers explorer
-
Unconstrained Robust Online Convex Optimization
New algorithms achieve regret roughly ||u||G(sqrt(T)+k) under adversarial gradient corruption in unconstrained online convex optimization, with a matching lower bound when G is known.