Exponential Convergence of Augmented Primal-dual Gradient Algorithms for Partially Strongly Convex Functions

TL;DR

This paper proves that augmented primal-dual gradient algorithms can achieve exponential convergence for functions that are strongly convex only in a subspace, extending to distributed optimization with relaxed conditions.

Contribution

It establishes global exponential convergence for augmented primal-dual algorithms under partial strong convexity and applies the results to distributed optimization with relaxed assumptions.

Findings

Achieves exponential convergence with partially strongly convex functions.

Extends convergence results to distributed optimization with relaxed conditions.

Unifies existing centralized and distributed algorithms under a common framework.

Abstract

We show that the augmented primal-dual gradient algorithms can achieve global exponential convergence with partially strongly convex functions. In particular, the objective function only needs to be strongly convex in the subspace satisfying the equality constraint and can be generally convex elsewhere, provided the global Lipschitz condition for the gradient is satisfied. This condition implies that states outside the equality subspace will converge towards it exponentially fast. The analysis is then applied to distributed optimization, where the partially strong convexity can be relaxed to the restricted secant inequality condition, which is not necessarily convex. This work unifies global exponential convergence results for some existing centralized and distributed algorithms.