Implement compatibility with non-deterministic games

snowfrogdev / macao

A general purpose game playing A.I. framework based on the Monte Carlo tree search algorithm.

MIT License

23 stars 4 forks source link

Implement compatibility with non-deterministic games #8

Open snowfrogdev opened 6 years ago

snowfrogdev commented 6 years ago

Some games, have elements of chance or randomness to them. Modify the MCTS algorithm to be compatible with such games.

Resources:

flesler commented 5 years ago

I really like this project, specially the simplicity of it. Hope you can make it able to handle more cases

snowfrogdev commented 5 years ago

@flesler Thanks. Have you had a chance to give it a try? If so, with what kind of game? How did it perform? Is there a game in particular you would like Macao to work with?

flesler commented 5 years ago

I've been experimenting with NN and Neuroevolution for AI and discovered this library (not related to ML I know). So I want to try it for some cases, the first one will probably be a snake that is very good at it (so no 2nd player) by looking many moves ahead and deciding what to do. Haven't tried it yet, will post as soon as I do