Skip to content

ShivankYadav/Q-Learning-on-Blackjack

Repository files navigation

Q-Learning-on-Blackjack

Teaching an Agent to play Blackjack using Q-Learning. The code is explained in the Monte_Carlo.ipynb. You guys are welcome to imporve the hyperparameters or even the algorithm for better performance. I have also provided a presentation and a report to explain Monte Carlo Control Algorithms.

Environment Description

https://github.com/openai/gym/blob/master/gym/envs/toy_text/blackjack.py

Install requirements

Simply execute this on your shell: $pip install -r requirements.txt For visualizations, I was not able to install basemap using pip. So I used conda install -c anaconda basemap

Algorithm used:

"Algorithm_image"