Boardgames: Chess & Checkers
These Chess and Checkers games were both developed while I was working for Natural Selection, Inc. through separate National Science Foundation (NSF) SBIR grants with the focus being the development of Neural Networks (NN) optimized through the use of Evolutionary Computation (EC) which means that the games were adaptive and self-learning. To clarify, the initial rules were programmed and a population of game-playing agents were created, each with its own neural network and these individual agents were evolved through the use of Darwinian concepts of Natural Selection (survival of the fittest) such that, within the simulation, only the best performers survived on to the next generation of individuals which were mutated from the surviving parents. Simple explanation: the games were programmed to learn how to play chess and checkers on their own and reached the rating of Expert for Checkers and Grand Master 2650 for Chess. As for myself, the programmer, I am a poor chess and checkers player and the programs learned to play at these ratings on their own, thus proving how well the system works. "Chess with an Attitude!" was shown to the famous chess player Gary Kasparov who seemed to enjoy playing it for a bit. I even had a patent awarded for the game based-on the design as a system.
A Self-Learning Evolutionary Chess Program
Boston’s Museum Of Science Features "Evolutionary Checkers Starring Blondie24" by Digenetics, Inc.