Tic-Tac-Toe with Reinforcement Learning
This is a repository for training an AI agent to play Tic-tac-toe using reinforcement learning. Both the SARSA and Q-learning RL algorithms are implemented. A user may teach the agent himself by playing the game for a couple of rounds, or he may apply an automated teacher agent.
Code Structure
Source Code
The directory tictactoe contains the core source code for this project.
There are 3 main source code files:
- game.py
- agent.py
- teacher.py
I have implemented two RL agents that learn to play the game of tic-tac-toe: one follows the SARSA algorithm, and the other follows Q-learning. These agents are trained by a teacher agent that knows the optimal strategy; however, the teacher only follows this strategy with a given probability p at each turn. The rest of the time this teacher chooses randomly from the moves that are available, so that the agents are able to win on occasion and learn from these wins. To initialize the learning agent Q values, I make use of default dictionaries with default values of 0 such that the value for every state-action pair is initialized to 0.
The Q-learning and SARSA agents are implemented in agent.py.
Each of the two learning agents inherit from a parent learner class; the key difference between the two is their Q-value update function.
The Teacher agent is implemented in teacher.py.
The teacher knows the optimal policy for each state presented; however, this agent only takes the optimal choice with a set probability.
In game.py, the main game class is found.
The Game class holds the state of each particular game instance, and it contains the majority of the main game functionality.
The main game loop can be found in the class's function playGame().
Game Script
To play the game (see "Running the Program" below for instructions) you will use the script called play.py.
This script is organized as follows.
The GameLearner class holds the state of the current game sequence, which will continue until the player choses to stop or the teacher has finished the designated number of episodes.
Running the Program
THESE INSTRUCTIONS ARE OUTDATED. They will be updated soon.
Train a new agent manually
To initialize a new agent and begin a new game loop, simply run:
python play.py -a q (Q-learner)
python play.py -a s (Sarsa-learner)
This will initialize the game and allow you to train the agent manually by playing against the agent repeatedly. Be careful, however, as initializing a new agent will delete the state of the previous agents that were stored for the selected agent type. In the process of playing, you will be storing the new agent state with each game iteration.
Train a new agent automatically (via teacher agent)
To initialize a new RL agent and train it automatically with a teacher agent, use the flag '-t' followed by the number of game iterations you would like to train for:
python play.py -a q -t 5000 (Q-learner)
python play.py -a s -t 5000 (Sarsa-learner)
Again, be careful as this will overwrite previously-existing agents.
Load a trained agent
To load existing agents and play against them, run:
python play.py -a q -l (Q-learner)
python play.py -a s -l (Sarsa-learner)
I have trained an instance of each the Q-learner and Sarsa-learner agents and pickled them into .pkl files that are included here. These agents were trained by a teacher of level 0.9 for 100000 episodes, and they have learned to play considerably well. You can make use of these if you like, but they will be overwritten if you have initialized new agents.
Load a trained agent and train it further
You can train existing agents further by loading them and teaching them, via a combination of '-t' and '-l':
python play.py -a q -l -t 5000 (Q-learner)
python play.py -a s -l -t 5000 (Sarsa-learner)
Load a trained agent and view reward history plot
Finally, to load a stored agent and view a plot of its cumulative reward history, use '-l' in combination with '-p':
python play.py -a q -l -p (Q-learner)
python play.py -a s -l -p (Sarsa-learner)