Deep Reinforcement Learning with PyTorch

• 1. Dynamic Programming (Update : 13. 2. 2019)

  1. Conditional GAN

  

  2. Policy Iteration & Value Iteration

• 2. Value Based Methods (Update : 17. 2. 2019)

  1. Vanilla DQN

  2. PDD DQN

  

• 3. Policy Based Methods (Update : 23. 2. 2019)

  1. A2C

  2. PPO

  

• 4. Off-policy Policy Based Methods (Update : 10. 3. 2019)

  1. SAC

  2. SIL ( not with A2C, PPO but SAC)

  

• 5. Exploration Techniques (Update : 16. 3. 2019)

  1. Thompson sampling with MCDO
  2. RND

  Breakout with only intrinsic rewards

  

• 6. Uncertainty in RL (Update : 24. 3. 2019)

  1. Categorical DQN(C51)
  2. QR-DQN
  3. Implicit Quantile Networks

  

• 7. Imitation Learning (Update : 30. 3. 2019)

  1. GAIL

• 8. Multi-Agent RL (Update : 4. 4. 2019)

  1. Upper Confidence Bounds for Tree(UCT)
  2. Counterfactual Hedge