Snake minigame-style with fixed goal, random initial position per episode OpenAI gym. Single Q learning agent.
