Implement a few reinforcement learning algorithms that can play Settlers of Catan!
To run this code, you only need to run 7_hex_Catan.ipnyb - the 7 hexagon version of Settlers of Catan
- the code should be set up to be able to run from top to bottom, except for the deprecated code in the bottom of the notebook
- there is a single line which breaks to code after the Sarsa(0) algorithm runs, as to not lose progress through variable redeclarations in the semi-gradient Sarsa(0) algorithm.
The other files are images and gifs which illustrate the implementation of the project and the actual game board in selected episodes. In the future, I hope to expend this project to implement the full Catan game using only non-tabular methods.