{"repo":"ai-boson/mcts","free":true,"listed":false,"github":"https://github.com/ai-boson/mcts","clone":"git clone https://github.com/ai-boson/mcts.git","description":"MCTS algorithm tutorial and it's explanation with code. Application of MCTS to create A.I for simple game.","language":"Ruby","stars":34,"topics":["mcts","mcts-algorithm","game-development","artificial-intelligence","ml"],"license":"MIT","category":"game-templates","readme_excerpt":"Monte Carlo Tree Search (MCTS) algorithm tutorial with Python code. Application of MCTS to create A.I for simple game. In this tutorial we will be explaining the Monte Carlo Tree Search algorithm and each part of the code. Recently we applied MCTS to develop our game. The code is general and only assumes familarity with basic Python. We have explained it with respect to our game. If you want to use it for your project or game, you will have to slightly modify the functions which I have mentioned below. Here is the playstore link to the game: [Sudo Tic Tac Toe][jekyll-talk]. Rules for our game(mode 1) are as follows: 1. The game is played on a 9 by 9 grid like Sudoku. 2. This big 9 by 9 grid is divided into 9 smaller 3 by 3 grids (local board). 3. Aim of the game is to win any one local board of the 9 available. 4. Your move determines in which local board A.I has to make a move and viceversa. 5. For example you make a move in position 1 of local board number 5. This will force the A.I to make a move in local board number 1. 6. Rules of normal Tic Tac Toe are applied to local board. Why MCTS? As you would have seen this game has a very high branching factor. For the first move the entire board is empty. So there are 81 empty spots. For the first turn it has 81 possible moves. For the second turn by applying rule 4 it has 8 or 9 possible moves. For the first 2 moves this results in 81 9 = 729 possible combinations. Thus the number of possible combinations increases as the game ","default_branch":null,"files":null,"tree":[],"storefront":"/r/ai-boson","claimed":false,"request_supported":{"post":"https://gitbuyer.com/r/ai-boson/mcts/request-supported","requests":0},"note":"indexed from public GitHub; nothing is for sale on this page. Clone it from GitHub. Paid listings live at /search."}