how to use

Set this root dir as PYTHONPATH.

export PYTHONPATH=$(pwd)

Contents

predl

predl contains some algorithms without Deep Learning. Here, PyTorch is just numpy-alternative.

• lookup table

• n-step SARSA

• SARSA(λ)

• Q-Learning

• SARSA

• DynaQ (Model Free Approach)

• function approximation

• Q-learning

• SARSA

• policy gradient

• REINFORCE

• REINFORCE with baseline

• actor-critic

dl

dl contains some algorithms using DL.

• Deep Q-Network

  cd dl/dqn
  python exec.py [--env Pong ...] # `python exec.py -h` for help
  ---
  tensorboard log_dir runs

Dependencies

• gym

• gym requires a lot depends on your environment thus read the official document before using pip.

• PyTorch

• matplotlib

• tensorboard-pytorch

• pip install tensorboardX

References

• UCL Course on RL by David Silver, DeepMind.

• I mainly learned RL in this lecture.

• Lecture videos are available on YouTube

• Richard S. Sutton and Andrew G. Barto Reinforcement Learning: An Introduction. 2017.

• online draft is available

• Introduction to reinforcement learning and OpenAI Gym by Justin Francis

• Good introduction to OpenAI Gym

• 牧野貴樹ほか.これからの強化学習.2016.

• 基礎から応用まで扱っています