activity
20242026
collaborators
Showing math.OCShow all

8 papers · 1 filter

math.OC2026

Negative Stepsizes Make Gradient-Descent-Ascent Converge

Henry Shugart, Jason M. Altschuler

Efficient computation of min-max problems is a central question in optimization, learning, games, and control. Arguably the most natural algorithm is gradient-descent-ascent (GDA).…

math.OC2026

Stepsize Hedging: an Alternative Mechanism for Accelerating Gradient Descent

Jason M. Altschuler, Pablo A. Parrilo

Can gradient descent be accelerated by just choosing better stepsizes? Surprisingly, the answer is yes. This short expository article provides an accessible introduction to this ph…

math.OC2026

Acceleration by Random Stepsizes: Hedging, Equalization, and the Arcsine Stepsize Schedule

Jason M. Altschuler, Pablo A. Parrilo

We show that for separable convex optimization, random stepsizes fully accelerate Gradient Descent. Specifically, using inverse stepsizes i.i.d. from the Arcsine distribution impro…

math.OC2026

Negative Momentum for Convex-Concave Optimization

Henry Shugart, Shuyi Wang, Jason M. Altschuler

This paper revisits momentum in the context of min-max optimization. Momentum is a celebrated mechanism for accelerating gradient dynamics in settings like convex minimization, but…

math.OC2025

Min-Max Optimization Is Strictly Easier Than Variational Inequalities

Henry Shugart, Jason M. Altschuler

Classically, a mainstream approach for solving a convex-concave min-max problem is to instead solve the variational inequality problem arising from its first-order optimality condi…

math.OC2025

Optimized methods for composite optimization: a reduction perspective

Jinho Bok, Jason M. Altschuler

Recent advances in convex optimization have leveraged computer-assisted proofs to develop optimized first-order methods that improve over classical algorithms. However, each optimi…