Proposes federated adaptive optimizers (FedAdagrad, FedAdam, FedYogi) with convergence analysis for non-convex objectives under data heterogeneity and reports empirical gains over FedAvg.
Linear Coupling: An Ultimate Unification of Gradient and Mirror Descent
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
abstract
First-order methods play a central role in large-scale machine learning. Even though many variations exist, each suited to a particular problem, almost all such methods fundamentally rely on two types of algorithmic steps: gradient descent, which yields primal progress, and mirror descent, which yields dual progress. We observe that the performances of gradient and mirror descent are complementary, so that faster algorithms can be designed by LINEARLY COUPLING the two. We show how to reconstruct Nesterov's accelerated gradient methods using linear coupling, which gives a cleaner interpretation than Nesterov's original proofs. We also discuss the power of linear coupling by extending it to many other settings that Nesterov's methods cannot apply to.
verdicts
UNVERDICTED 2representative citing papers
APAPC integrates Nesterov acceleration into primal-dual forward-backward schemes by exploiting dual strong convexity to achieve optimal sublinear and accelerated linear convergence rates.
citing papers explorer
-
Adaptive Federated Optimization
Proposes federated adaptive optimizers (FedAdagrad, FedAdam, FedYogi) with convergence analysis for non-convex objectives under data heterogeneity and reports empirical gains over FedAvg.
-
A Nesterov-Accelerated Primal-Dual Splitting Algorithm for Convex Nonsmooth Optimization
APAPC integrates Nesterov acceleration into primal-dual forward-backward schemes by exploiting dual strong convexity to achieve optimal sublinear and accelerated linear convergence rates.