Regularized extragradient method for searching for an equilibrium point in two-person saddle-point games

Abstract

A two person saddle point game with approximately given input data is examined. Since, in games of this type, the search for an equilibrium point is unstable with respect to perturbations in the input data, two variants of the regularized extragradient method are proposed. Their convergence is analyzed, and a regularizing operator is constructed.