Understanding Modern Techniques in Optimization: Frank-Wolfe, Nesterov's Momentum, and Polyak's Momentum

TL;DR

This paper introduces a modular game-theoretic framework for analyzing and developing optimization algorithms like Frank-Wolfe and Nesterov's acceleration, and explores Polyak's momentum for acceleration and saddle point escape in non-convex problems.

Contribution

It presents a unified game-based framework for convex optimization algorithms, develops new Frank-Wolfe-like methods, and provides modular analysis of Polyak's momentum in modern non-convex optimization.

Findings

Game-theoretic framework recovers existing algorithms

New Frank-Wolfe-like algorithms for specific constraint sets

Polyak's momentum accelerates convergence and aids saddle point escape

Abstract

In the first part of this dissertation research, we develop a modular framework that can serve as a recipe for constructing and analyzing iterative algorithms for convex optimization. Specifically, our work casts optimization as iteratively playing a two-player zero-sum game. Many existing optimization algorithms including Frank-Wolfe and Nesterov's acceleration methods can be recovered from the game by pitting two online learners with appropriate strategies against each other. Furthermore, the sum of the weighted average regrets of the players in the game implies the convergence rate. As a result, our approach provides simple alternative proofs to these algorithms. Moreover, we demonstrate that our approach of optimization as iteratively playing a game leads to three new fast Frank-Wolfe-like algorithms for some constraint sets, which further shows that our framework is indeed generic, modular, and easy-to-use. In the second part, we develop a modular analysis of provable acceleration via Polyak's momentum for certain problems, which include solving the classical strongly quadratic convex problems, training a wide ReLU network under the neural tangent kernel regime, and training a deep linear network with an orthogonal initialization. We develop a meta theorem and show that when applying Polyak's momentum for these problems, the induced dynamics exhibit a form where we can directly apply our meta theorem. In the last part of the dissertation, we show another advantage of the use of Polyak's momentum -- it facilitates fast saddle point escape in smooth non-convex optimization. This result, together with those of the second part, sheds new light on Polyak's momentum in modern non-convex optimization and deep learning.