/Practical-RL-coursera

Primary LanguageJupyter Notebook

Reinforcement Learning algorithms

My Implementations of the Reinforcement Learning algorithms from the Practical_Rl course from Coursera using TensorFlow

In the next few commits, i'll add implementations in PyTorch as well.

The following algorithms have been implemented in these notebooks:
Week 1:
  1. Introduction to TensorFlow
  2. Introduction to Gym
  3. Cross-entropy method
  4. Deep cross entropy method
Week 2:
  1. Markov Decision Process
Week 3:
  1. Q-Learning
  2. Expected value SARSA
  3. Q-Learning using experience replay
Week 4:
  1. Approximate Q-Learning
  2. DQN on Atari: Breakout.
Week 5:
  1. REINFORCE
  2. Asynchronous Actor Critic method (A3C)
Week 6:
  1. Multi Arm bandits (including different approximation methods)
  2. Monte Carlo Tree Search

Research Papers: