1. Paper title: Genetic Algorithms for Evolving Computer Chess Programs. ------------------------------------------------------------------------------- 2. Author contact details: Omid E. David Department of Computer Science, Bar-Ilan University, Ramat-Gan 52900, Israel mail@omiddavid.com H. Jaap van den Herik Tilburg Center for Cognition and Communication, Tilburg University, Tilburg 5037 AB, The Netherlands h.j.vdnherik@uvt.nl Moshe Koppel Department of Computer Science, Bar-Ilan University, Ramat-Gan 52900, Israel koppel@cs.biu.ac.il Nathan S. Netanyahu Department of Computer Science, Bar-Ilan University, Ramat-Gan 52900, Israel nathan@cs.biu.ac.il ------------------------------------------------------------------------------- 3. Corresponding author: Omid E. David mail@omiddavid.com ------------------------------------------------------------------------------- 4. Paper abstract: This paper demonstrates the use of genetic algorithms for evolving (1) a grandmaster-level evaluation function and (2) a search mechanism for a chess program, the parameter values of which are initialized randomly. The evaluation function of the program is evolved by learning from databases of (human) grandmaster games. At first the organisms are evolved to mimic the behavior of human grandmasters, and then these organisms are further improved upon by means of coevolution. The search mechanism is evolved by learning from tactical test suites. Our results show that the evolved program outperforms a two-time World Computer Chess Champion, and is on a par with other leading computer chess programs. ------------------------------------------------------------------------------- 5. Relevant criteria: B, G, H ------------------------------------------------------------------------------- 6. Statement: In this paper, we present a novel approach for evolving all the key components of a chess program (evaluation function + search mechanism) from randomly initialized values using genetic algorithms. This is achieved using only a database of human grandmaster games as training data. The evolved program is on a par with top commercial computer chess programs and outperforms a two-time World Computer Chess Champion. Performing on a par with top commercial computer chess programs puts the evolved program at a playing level higher than that of human grandmaster chess players. This paper provides the first methodology for automatic learning of all the key parameters of a computer chess program (search and evaluation components), and achieving state-of-the-art performance. ------------------------------------------------------------------------------- 7. Citation: Omid E. David, H. Jaap van den Herik, Moshe Koppel, and Nathan S. Netanyahu (2013): Genetic Algorithms for Evolving Computer Chess Programs. IEEE Transactions on Evolutionary Computation. doi:10.1109/TEVC.2013.2285111. In press. http://dx.doi.org/10.1109/TEVC.2013.2285111 Note: Please use the following link instead of posting the pdf on the website: http://www.oedavid.com/pubs~/ga-chess.pdf ------------------------------------------------------------------------------- 8. Any prize money, if any, is to be divided equally among the co-authors. ------------------------------------------------------------------------------- 9. "Best" statement: The enormously complex game of chess, referred to as "the touchstone of the intellect" by Goethe, has always been one of the main battlegrounds of man versus machine. John McCarthy refers to chess as the "Drosophila of AI". Computer chess, while being one of the most researched fields within AI, has not lent itself to the successful application of conventional learning methods, due to its enormous complexity. In this paper we demonstrate how GA can be used to successfully evolve all components of the chess program, and achieve super grandmaster-level performance. ------------------------------------------------------------------------------- 10. Method used: Genetic Algorithms (GA).