Comparing Reinforcement Learning Algorithms in the Moonlander Environment