Solve the optimized path of a maze using Q-Learning of the Reinforcement Learning and epsilon-greedy.