Programming Chess #5 - Really Thinking Like A Human This Time
Neural Networks are basically a computer simulation of the human brain. Don't worry- robots won't take over the world just yet. These things are incredibly limited, and about the best they can do right now is factor prime numbers.
Beyond that, I'm no specialist- check out http://colinfahey.com/neural_network_with_back_propagation_learning/neural_network_with_back_propagation_learning_en.html for a more in-depth look at these things, including applications to complex board games. Well, tic-tac-toe.
So, I suppose theoretically you could build one of these to play chess. If anyone knows exactly how impractical that would be, I'd love to hear it.
Next time: back into the main series with- positional play?!?