Efficient Projection-Free Algorithms for Saddle Point Problems
Abstract
The Frank-Wolfe algorithm is a classic method for constrained optimization problems. It has recently been popular in many machine learning applications because its projection-free property leads to more efficient iterations. In this paper, we study projection-free algorithms for convex-strongly-concave saddle point problems with complicated constraints. Our method combines Conditional Gradient Sliding with Mirror-Prox and shows that it only requires gradient evaluations and linear optimizations in the batch setting. We also extend our method to the stochastic setting and propose first stochastic projection-free algorithms for saddle point problems. Experimental results demonstrate the effectiveness of our algorithms and verify our theoretical guarantees.
Cite
@article{arxiv.2010.11737,
title = {Efficient Projection-Free Algorithms for Saddle Point Problems},
author = {Cheng Chen and Luo Luo and Weinan Zhang and Yong Yu},
journal= {arXiv preprint arXiv:2010.11737},
year = {2020}
}