Accelerated GRAAL is the first adaptive first-order method that proves near-optimal accelerated complexity for convex L-smooth and (L0,L1)-smooth functions with geometric stepsize growth.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
math.OC 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Nesterov Finds GRAAL: Optimal and Adaptive Gradient Method for Convex Optimization
Accelerated GRAAL is the first adaptive first-order method that proves near-optimal accelerated complexity for convex L-smooth and (L0,L1)-smooth functions with geometric stepsize growth.