Accelerated Methods for Riemannian Min-Max Optimization Ensuring Bounded Geometric Penalties

2023/05/25 by David Martínez-Rubio, Christophe Roux, Martínez-Rubio, David +5
Mathematics · #Advanced Optimization Algorithms Research #FOS: Computer and information sciences #FOS: Mathematics #Machine Learning (cs.LG) #Mathematical Biology Tumor Growth #Optimization and Control (math.OC)

paper · pdf · doi:10.48550/arxiv.2305.16186

openalex publication_date 2023/05/25 · openalex created_date 2023/05/27 · openalex updated_date 2026/07/28

Abstract

In this work, we study optimization problems of the form min_x max_y f(x, y), where f(x, y) is defined on a product Riemannian manifold M × N and is μ_x-strongly geodesically convex (g-convex) in x and μ_y-strongly g-concave in y, for μ_x, μ_y ≥ 0. We design accelerated methods when f is (L_x, L_y, L_xy)-smooth and M, N are Hadamard. To that aim we introduce new g-convex optimization results, of independent interest: we show global linear convergence for metric-projected Riemannian gradient descent and improve existing accelerated methods by reducing geometric constants. Additionally, we complete the analysis of two previous works applying to the Riemannian min-max case by removing an assumption about iterates staying in a pre-specified compact set.

Related

• Accelerated Riemannian Optimization: Handling Constraints with a Prox to Bound Geometric Penalties
• First-Order Algorithms for Min-Max Optimization in Geodesic Metric Spaces
• Global Riemannian Acceleration in Hyperbolic and Spherical Spaces
• Momentum Improves Optimization on Riemannian Manifolds
• First-order Methods for Geodesically Convex Optimization