/Uno-Game-with-Reinforcement-Learning-Algorithm

Primary LanguagePython

Uno-Game-with-Reinforcement-Learning-Algorithm

This repository creates an entire Uno game and implements three Reinforcement Learning Algorithms:

  1. Monte Carlo Algorithm
  2. SARSA
  3. Q-Learning

It makes the three algorithms play each other in the Main.py code to check which performs the best in the long run in this game.

Steps to run the code:

  1. Clone the repository.
  2. Run the Uno_Game.py if you want to understand how the game works and the function of every class. It provides a detailed step-by-step output of the entire game.
  3. Run the Monte_Carlo.py, SARSA.py, and Q_Learning.py to check the algorithm's performance against a computer without any algorithm and with a purely randomized approach. The algorithms play with Player 1 while Player 2 is the computer.
  4. Run the Main.py for the entire training of all three algorithms and a final match results when the three play against each other.