Stichworte: reinforcing successive approximations