Web interfaces in HTML, css and javascript for
- Tic-Tac-Toe
- 4-in-a-row
- Tetris
This repository is a collection of AI algorithm to play games, you can play against the AIs at the github pages https://andreariba.github.io/games-ai. Currently, the implemented algorithms include
| Game | minimax | Monte Carlo Tree Search | AlphaZero-like NN | Deep Q-learning |
|---|---|---|---|---|
| Tic-Tac-Toe | ✔️ full depth | in progress | ✔️ | not planned |
| 4-in-a-row | ✔️ depth=4 | in progress | ✔️ | not planned |
| Tetris | not planned | in progress | in progress | in progress |
The AlphaZero-like AI are still in a preliminary state, working on improving them.
Since the simplicity of the algorithm, I strongly recommend to read the Wiki page relative to the minimax, https://en.wikipedia.org/wiki/Minimax
-
Brief overview on the idea's behind AlphaZero, https://medium.com/mlearning-ai/mastering-tictactoe-with-alphazero-cc28998bf36c
-
Nicely written implementation of AlphaZero for the Connect 2 game by Josh Varty, instructive and easy to follow, https://github.com/JoshVarty/AlphaZeroSimple.
-
The original Nature's paper from AlphaZero to get into some of the details about the structure of the network and the parameters for the training, https://www.nature.com/articles/nature24270.
-
David Foster's Cheat sheet, https://medium.com/applied-data-science/alphago-zero-explained-in-one-diagram-365f5abf67e0.
