RL adalah cabang ML yang fokus pada sequential decision making.
Framework
Agent -> action -> Environment -> state + reward
^ |
+--------------------------------------+
Key Algorithms
| Algorithm | Type | Key Feature |
|---|---|---|
| DQN | Value-based | Experience replay + target network |
| PPO | Policy gradient | Clipped surrogate objective |
| SAC | Actor-critic | Maximum entropy, continuous action |
| TD3 | Actor-critic | Twin critics, target smoothing |
| AlphaZero | Self-play | MCTS + neural network |
Applications
- Robotics: manipulation, locomotion, grasping
- Game AI: AlphaGo, Dota 2
- Security: adversarial RL for penetration testing