Q-learning, policy gradient, actor-critic, and multi-agent reinforcement learning.