Accelerated Primal-Dual Proximal Gradient Splitting Methods for Convex-Concave Saddle-Point Problems

TL;DR

This paper introduces new accelerated primal-dual proximal gradient splitting methods for convex-concave saddle-point problems, achieving optimal nonergodic convergence rates with adaptive Bregman divergences and velocity correction techniques.

Contribution

It develops a novel accelerated primal-dual hybrid gradient flow with Bregman divergences, providing flexible, adaptive algorithms with provable optimal convergence for saddle-point problems.

Findings

Proposed methods achieve optimal nonergodic convergence rates.

The algorithms are flexible and adapt to various convexity and metric settings.

Theoretical analysis confirms exponential decay of Lyapunov functions.

Abstract

In this paper, based a novel primal-dual dynamical model with adaptive scaling parameters and Bregman divergences, we propose new accelerated primal-dual proximal gradient splitting methods for solving bilinear saddle-point problems with provable optimal nonergodic convergence rates. For the first, using the spectral analysis, we show that a naive extension of acceleration model for unconstrained optimization problems to a quadratic game is unstable. Motivated by this, we present an accelerated primal-dual hybrid gradient (APDHG) flow which combines acceleration with careful velocity correction. To work with non-Euclidean distances, we also equip our APDHG model with general Bregman divergences and prove the exponential decay of a Lyapunov function. Then, new primal-dual splitting methods are developed based on proper semi-implicit Euler schemes of the continuous model, and the theoretical convergence rates are nonergodic and optimal with respect to the matrix norms,\, Lipschitz constants and convexity parameters. Thanks to the primal and dual scaling parameters, both the algorithm designing and convergence analysis cover automatically the convex and (partially) strongly convex objectives. Moreover, the use of Bregman divergences not only unifies the standard Euclidean distances and general cases in an elegant way, but also makes our methods more flexible and adaptive to problem-dependent metrics.