Slides YouTube

• Policy based reinforcement learning
  • Stochastic policy
  • Policy Search
• Finite Difference Policy Gradient
• Monte-Carlo Policy Gradient
  • Score function
  • Policy gradient theorem
  • REINFORCE
• Actor-Critic Methods
  • Compatible function approximation
  • Variance reduction via baseline
  • Advantage Function Critic
  • Eligibility Traces
  • Natural Policy Gradient

• State of the art and applications

• Not yet covered
  • Deep Deterministic Policy Gradient(DDPG)
  • Asynchronous Advantage Actor-Critic Algorithm (A3C)
  • Trust Region Policy Optimization(TRPO), Proximal Policy Optimization(PPO)
  • Soft Actor-Critic