Question

I have an AI that is good at playing Connect 4 (using minimax). Now I want to use some machine learning algorithm to learn from this AI that I have, and I would like to do that by just letting them play against each other.

What algorithm would be good for this, and how would I train it? If someone could just name a way of doing this I can easily Google it by my self. But right now I don't know what to Google...

Was it helpful?

Solution

You could definitely use a neural network to do this. Since it can be hard to find the right amount of input and output nodes and all the weights, I recommend using evolutionary computation techniques (such as a genetic algorithm) to do this.

Hope this helps. Cheers!

OTHER TIPS

Connect Four is a solved game, meaning that there is a strategy that will always allow the player who goes first to win. You could try to do a machine learning approach, but it would pointless except as an exercise.

You can read how Victor Allis used an expert system to find the winning strategy in his master's thesis (pdf).

Licensed under: CC-BY-SA with attribution
Not affiliated with StackOverflow
scroll top