One-sided Frank-Wolfe algorithms for saddle problems

TL;DR

This paper introduces one-sided Frank-Wolfe algorithms for convex-concave saddle problems, achieving improved convergence rates and efficiency, with applications in machine learning and computer vision.

Contribution

It develops novel primal-dual algorithms combining Frank-Wolfe and proximal methods, with proven faster convergence rates for specific saddle-point problems.

Findings

Achieves $O(1/n^2)$ convergence rate for dual objective with linear constraints.

Provides $O(1/n^2)$ bounds on primal and infeasibility gaps in constrained optimization.

Demonstrates improved convergence rates over existing methods for saddle-point problems.

Abstract

We study a class of convex-concave saddle-point problems of the form $\min_x\max_y \langle Kx,y\rangle+f_{\cal{P}}(x)-h^\ast(y)$ where $K$ is a linear operator, $f_{\cal{P}}$ is the sum of a convex function $f$ with a Lipschitz-continuous gradient and the indicator function of a bounded convex polytope $\cal{P}$, and $h^\ast$ is a convex (possibly nonsmooth) function. Such problem arises, for example, as a Lagrangian relaxation of various discrete optimization problems. Our main assumptions are the existence of an efficient linear minimization oracle ($lmo$) for $f_{\cal{P}}$ and an efficient proximal map for $h^*$ which motivate the solution via a blend of proximal primal-dual algorithms and Frank-Wolfe algorithms. In case $h^*$ is the indicator function of a linear constraint and function $f$ is quadratic, we show a $O(1/n^2)$ convergence rate on the dual objective, requiring $O(n \log n)$ calls of $lmo$. If the problem comes from the constrained optimization problem $\min_{x\in\mathbb R^d}\{f_{\cal{P}}(x)\:|\:Ax-b=0\}$ then we additionally get bound $O(1/n^2)$ both on the primal gap and on the infeasibility gap. In the most general case, we show a $O(1/n)$ convergence rate of the primal-dual gap again requiring $O(n\log n)$ calls of $lmo$. To the best of our knowledge, this improves on the known convergence rates for the considered class of saddle-point problems. We show applications to labeling problems frequently appearing in machine learning and computer vision.