Аннотация:Quadratic functional minimization problems with constraints in Hilbert space are considered.
The feasible set is either a polyhedron or an ellipsoid. The dependence of the approximate solution accuracy on error level in initial data is investigated.