The regularized constraint aggregation method for a problem of semi-infinite optimization (Q1571159)

A regularized constraint aggregation method for convex problems of semi-infinite optimization is considered. The idea of the method is to simplify the structure of the feasible set by using, instead of the initial set, an extended set formed by convolution of the constraints of the original set. When the number of constraints is infinite, as a rule, the norm of the residual can be evaluated only approximately. This is taken into account in analyzing the convergence of the method.