Reinforcement Learning Terminology Part 2

Reinforcement Learning Terminology Part 2

Several important terms in Reinforcement Learning are explained in this video, including Intrinsic and Extrinsic rewards, reward shaping/engineering, reward hacking, model-free vs. model-based RL, Critic and Actor, Advantage function, offline RL, hybrid RL, Hierarchical RL, Imitation learning, Inverse RL, safe RL, constrained RL, multi-agent (cooperative and competitive) RL, experience replay buffer, decay schedule for epsilon-greedy, Thompson sampling, Multi-armed bandit problems, Regret, and Upper Confidence Bound algorithm.