Combined Use of Reinforcement Learning and SimulatedAnnealing by Peter Stefan