Abstract:Riemannian convex optimization and minimax optimization have recently drawn considerable attention. Their appeal lies in their capacity to adeptly manage the non-convexity of the objective function as well as constraints inherent in the feasible set in the Euclidean sense. In this work, we delve into monotone Riemannian Variational Inequality Problems (RVIPs), which encompass both Riemannian convex optimization and minimax optimization as particular cases. In the context of Euclidean space, it is established that the last-iterates of both the extragradient (EG) and past extragradient (PEG) methods converge to the solution of monotone variational inequality problems at a rate of $O\left(\frac{1}{\sqrt{T}}\right)$ (Cai et al., 2022). However, analogous behavior on Riemannian manifolds remains an open question. To bridge this gap, we introduce the Riemannian extragradient (REG) and Riemannian past extragradient (RPEG) methods. We demonstrate that both exhibit $O\left(\frac{1}{\sqrt{T}}\right)$ last-iterate convergence. Additionally, we show that the average-iterate convergence of both REG and RPEG is $O\left(\frac{1}{T}\right)$, aligning with observations in the Euclidean case (Mokhtari et al., 2020). These results are enabled by judiciously addressing the holonomy effect so that additional complications in Riemannian cases can be reduced and the Euclidean proof inspired by the performance estimation problem (PEP) technique or the sum-of-squares (SOS) technique can be applied again.

Extragradient-Type Methods with $\mathcal{O} (1/k)$ Last-Iterate Convergence Rates for Co-Hypomonotone Inclusions

Revisiting Extragradient-Type Methods -- Part 1: Generalizations and Sublinear Convergence Rates

On the Hypomonotone Class of Variational Inequalities

New Extragradient-Type Methods for Solving Variational Inequalities

Extragradient Type Methods for Riemannian Variational Inequality Problems

Double inertial extragradient algorithms for solving variational inequality problems with convergence analysis

Kantorovich-Type Convergence Criterion for Inexact Newton Methods

Extragradient methods with dual inertial steps for solving equilibrium problems

Primal-dual extrapolation methods for monotone inclusions under local Lipschitz continuity

Accelerated inertial subgradient extragradient algorithms with non-monotonic step sizes for equilibrium problems and fixed point problems

Viscosity extragradient algorithms with double inertial extrapolation steps for quasimonotone variational inequalities and fixed point problems

An Extragradient-Type Method for Generalized Equilibrium Problems Involving Strictly Pseudocontractive Mappings

On the convergence properties of non-Euclidean extragradient methods for variational inequalities with generalized monotone operators

Subgradient extragradient method with double inertial steps for quasi-monotone variational inequalities

Convergence of two-step inertial Tseng's extragradient methods for quasimonotone variational inequality problem

Extragradient method with feasible inexact projection to variational inequality problem

Accelerating Nesterov's Method for Strongly Convex Functions with Lipschitz Gradient

Convergence of Nonmonotone Proximal Gradient Methods under the Kurdyka-Lojasiewicz Property without a Global Lipschitz Assumption

Accelerated Variance-Reduced Forward-Reflected Methods for Root-Finding Problems

Modified Extragradient Methods for Solving Variational Inequalities