Introduction - If you have any usage issues, please Google them yourself
Reinforcement learning (Q-learning) has to solve this problem: an autonomous agent that can perceive the environment, and how to achieve the optimal action of its target through learning and selection.