In this repo you will find:
- Python implementation of most of the algorithms presented in Richard Sutton and Andrew Barto's well-known book Reinforcement Learning: An Introduction (second edtion). This is encapsulated in the IRL module.
- Python scripts for replicating most of the results presented in the book. These are collected in different folders corresponding to each chapter.
The code is written and organized with two major priorities:
- Pedagogy : to make algorithms and concepts as clear and as close to their form explained in the book as possible
- Modularity : to facilitate experimenting different algorithms with different problems
Dependencies
- Numpy
- Matplotlib (for visualization)