A priori and a posteriori error estimation for solutions of ill-posed problemsтезисы доклада