A regularized differential extraproximal method for finding an equilibrium in two-person saddle-point games

A two-person saddle-point game with approximate input data is considered. This problem is generally unstable with respect to the input data. To solve this problem, it is necessary to use specific methods called regularization methods. Three versions of the regularized differential extraproximal method are proposed and analyzed and their convergence is studied. A regularization operator is constructed.