installations:

pip install gym pip install numpy pip install matplotlib

to install all needed pip packages, run pip install -r requirements.txt

how blackjack works

CARD VALUES

• aces: 1 or 11
• all number cards: their values
• royalty: 10

RULES

• your goal is to beat the dealer and be as close to 21 as possible while not going over
• you win if: the dealer "busts" or you are closer to 21 then the dealer is
• you lose if: you "bust" or the dealer is closer to 21 than you are
• the dealer must: "stay" (as opposed to "hit", or get another card) if they are >= 17 in term of card value

this implements a q-learning algorithmic approach to train an agent into learning how to play blackjack using reinforcement learning

ERROR FIXING

• to be safe, when first initializing this project, run export PYTHONPATH='.' after you've cd'ed into the blackjack folder
• then, to simulate the reinforcement learning agent, run python3 agents/rl_agent.py
• to run the pygame, run python3 main.py