Irindul / TicTacToe

TicTacToe game with Q-Learner player. The AI is self-trained and became good after 500000 games.

TicTacToe

Simple TicTacToe game developed to learn Reinforcement Learning using Q-Learning.

The AI is basic and all was handmade, so obviously the performance is not great.

TODO

• Improve AI performance overall
• Support multiple policies (right now, we're using only epsilon-greedy)