Genetic algorithm behavior in the MAXSAT domain
Random Boolean Satisfiability function generators have recently been proposed as tools for studying genetic algorithm behavior. Yet MAXSAT problems exhibit extremely limited epistasis. Furthermore, all nonzero Walsh coefficients can be computed exactly for MAXSAT problems in polynomial time using only the clause information. This means the low order schema averages can be computed quickly and exactly for very large MAXSAT problems. But unless P=NP, this low order information cannot reliably lead to the global optimum, thus nontrivial MAXSAT problems must be deceptive.
Unable to display preview. Download preview PDF.
- 1.P. Cheeseman, B. Kanefsky, and W. M. Taylor. Where the really hard problems are. In Proc. Twelfth International Joint Conference on Artificial Intelligence, 1991.Google Scholar
- 3.J. Frank. A Study of Genetic Algorithms to Find Approximate Solutions to Hard 3CNF Problems. Golden West International Conference on Artificial Intelligence, 1994.Google Scholar
- 7.D. Mitchell, B. Selman, and H. Levesque. Hard and easy distribution of sat problems. In Proc. Tenth National Conference on Artificial Intelligence, San Jose, CA, 1992.Google Scholar
- 8.S. Rana, R. Heckendorn, and D. Whitley. A tractable walsh analysis of Sat and its implications for genetic algorithms. In Proc. Fifteenth National Conference on Artificial Intelligence, 1998.Google Scholar
- 9.C. Reeves and C. Wright. Epistasis in genetic algorithms: An experimental design perspective. In Larry Eshelman, editor, Proceedings of the Sixth International Conference on Genetic Algorithms, pages 217–224. Morgan Kaufmann, 1995.Google Scholar
- 10.B. Selman, H. Levesque, and D. Mitchell. A new method for solving hard satisfiability problems. In Proc. Tenth National Conference on Artificial Intelligence, pages 440–446, San Jose, CA, 1992.Google Scholar