Under convex (L0,L1)-smoothness, GD, NGD, Clip-GD, RCD and OrderRCD converge linearly while gradient norms stay above L0/L1, then revert to sublinear O(1/N) convergence.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
math.OC 1years
2024 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Linear Convergence Rate in Convex Setup is Possible! Gradient Descent Method Variants under $(L_0,L_1)$-Smoothness
Under convex (L0,L1)-smoothness, GD, NGD, Clip-GD, RCD and OrderRCD converge linearly while gradient norms stay above L0/L1, then revert to sublinear O(1/N) convergence.