Batched normalized SGD with momentum reaches the optimal heavy-tailed nonconvex rate without gradient clipping, and attains a slower but parameter-free rate when the tail index is unknown.
Martingale transforms
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
method 1
citation-polarity summary
fields
math.OC 1years
2024 1verdicts
CONDITIONAL 1roles
method 1polarities
use method 1representative citing papers
citing papers explorer
-
Nonconvex Stochastic Optimization under Heavy-Tailed Noises: Optimal Convergence without Gradient Clipping
Batched normalized SGD with momentum reaches the optimal heavy-tailed nonconvex rate without gradient clipping, and attains a slower but parameter-free rate when the tail index is unknown.