I already developed a framework for game of hex which chooses optimal choices against player based on Monte Carlo simulation. In order to evaluate the current moves in the tree, I used UCT and RAVE in combination with alpha AMAF.

I was wondering if I could mix the evolutionary methods (PSO, Genetic Algorithm, Ant Colony, etc) to enhance the algorithm.