Combined Use of Reinforcement Learning and SimulatedAnnealing door Peter Stefan