Variational Inequalities Overview

Principles

Definition

Variational inequalities is a class of mathematical problems or a framework concerned with finding a point within a given set that satisfies a specific inequality condition involving a given operator. This concept serves as a powerful and unifying approach for analyzing and solving diverse problems arising in optimization theory, equilibrium modeling, and the study of differential equations across various scientific and engineering disciplines.

Ontological type

Applications

Core Methods

Key Classes

Key developments

Key Figures

Functional-Analytic Foundations era

R. T. Rockafellar [1] is associated with the University of Washington [3] and Princeton University [4] in the Functional-Analytic Foundations era (1966–1988). R. T. Rockafellar [1] authored Monotone Operators and the Proximal Point Algorithm [7], which established a foundational link between monotone operator theory and proximal methods and thereby anchored the era's dual representations and convergence analyses. Jong‐Shi Pang [2] is associated with Johns Hopkins University [5] and the University of Wisconsin–Madison [6] in this era. His works, The Generalized Quasi-Variational Inequality Problem [8] and Asymmetric variational inequality problems over product sets: Applications and iterative methods [9], broadened the variational inequality framework to quasi-variational and asymmetric settings and spurred iterative solution methods central to the era.

Splitting and Decomposition Methods era

Dimitri P. Bertsekas [1] worked across the Massachusetts Institute of Technology [3] and the University of California, Santa Barbara [4] during the Splitting and Decomposition era. His 1992 article On the Douglas—Rachford splitting method and the proximal point algorithm for maximal monotone operators [6] introduced splitting strategies that underlie modular decomposition and provide convergence guarantees essential for large-scale variational inequalities. Paul Tseng [2] pursued research across the Massachusetts Institute of Technology [3] and the University of Washington [5] in this era. His splitting-based methods, including the Modified Forward-Backward Splitting Method for Maximal Monotone Mappings [7], the Applications of a Splitting Algorithm to Decomposition in Convex Programming and Variational Inequalities [8], and the Modified Projection-Type Methods for Monotone Variational Inequalities [9], provided modular, convergent approaches that advanced decomposition and parallelizable solvers during this era.

Acceleration and Projection-Extrapolation era

Aviv Gibali[1], affiliated with Technion – Israel Institute of Technology[3] and the University of Haifa[4], is a prominent figure in variational inequality theory during this era. His key contributions include the Subgradient Extragradient Method for Solving Variational Inequalities in Hilbert Space[7], Extensions of Korpelevich's extragradient method for the variational inequality problem in Euclidean space[9], and Algorithms for the Split Variational Inequality Problem[8], which propelled acceleration by providing explicit iteration-complexity bounds and enabling robust handling of monotone, nonsmooth, and split-structure problems. Simeon Reich[2], affiliated with University of California, Santa Barbara[5] and Carnegie Mellon University[6], is another leading figure in this era. His work on the Subgradient Extragradient Method for Solving Variational Inequalities in Hilbert Space[7], Algorithms for the Split Variational Inequality Problem[8], and Extensions of Korpelevich's extragradient method for the variational inequality problem in Euclidean space[9] contributed to unified projection-extrapolation frameworks and provided rigorous iteration-complexity analyses essential for acceleration in monotone and Lipschitz settings.