This TicTacToe app written in Swift was made within the scope of my graduation thesis.
There are two versions of the game. In each the AI is implemented using a different algorithm
- MinMax with Alpha-Beta Pruning
- Reinforcement Learning (Temporal Difference Algorithm)
The training method was to let 2 AI-s which use the same algorithm (TD) play against each other for 10,000 iterations. Below are presented the results of the AI against a human player after different number of training iterations.
| 1,000 iterations | 5,000 iterations | 10,000 iterations | |
|---|---|---|---|
| Human Wins | 5 | 7 | 0 |
| Draw | 5 | 3 | 10 |
| AI Wins | 0 | 0 | 0 |
Better results for the 1,000 iteration case are due to the human player requiring more time to find out the stategy to abuse.