No description, website, or topics provided.
Switch branches/tags
Nothing to show
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
__pycache__
logs
.gitignore
Blackjack- DQN (Only Hit or Stand).ipynb
Blackjack- Q Learning (Doubling Down).ipynb
Blackjack-Q Learning (Only Hit or Stand) 100 Lines.ipynb
Blackjack-Q Learning (Only Hit or Stand).ipynb
README.md
blackjackenv.py
optimal_policy.csv

README.md

Blackjack--Reinforcement-Learning

Teaching a bot how to play Blackjack using two techniques: Q-Learning and Deep Q-Learning.

The game used is OpenAI's gym environment. However, the blackjack game only consists of hitting and standing.

To fully obtain a working Blackjack bot, it would be necessary to add doubling down, splitting, and variation of bets to the game environment.

TODO

  1. Complete Double Down Implementation
  2. Complete Splitting Implentation
  3. Complete Varying Bet Implementation
  4. Understand why Q-Learner isn't learning doubling down.
  5. Understand why DQN isn't learning well.