This is a simple implementation of Q-learning for tic-tac-toe.
python3 game.py
The game agent will train for 1000 games using self-play, and then it will begin a game with the user.
The user may use numbers 0-8 as actions, corresponding to
0 | 1 | 2
-----------
3 | 4 | 5
-----------
6 | 7 | 8
or the letters q, w, e, a, s, d, z, x, c, which correspond to the same moves:
q | w | e
-----------
a | s | d
-----------
z | x | c
changes