Simple example of reinforcement learning using the "Sarsa" method with epsilon-greedy to solve the optimized path of a maze.