🧭 Phase 3 — Policy-Based Methods

Directly learn parameterized policies without discrete value tables.

Topics

• Policy Gradient Theorem
• REINFORCE Algorithm
• Variance Reduction (Baselines)
• Actor-Critic (A2C)

Mini Projects

• MountainCarContinuous-v0 (REINFORCE)
• CartPole (A2C)

📁 Source folder:
03-Policy-Based