Modelling exceptionally hard constraint satisfaction problems
Randomly-generated binary constraint satisfaction problems go through a phase transition as the constraint tightness varies. Loose constraints give an ‘easy-soluble’ region, where problems have many solutions and are almost always easy to solve. However, in this region, systematic search algorithms may occasionally encounter problems which are extremely expensive to solve. It has been suggested that in these cases, the first few instantiations made by the algorithm create an insoluble subproblem; an exhaustive search of the subproblem to prove its insolubility accounts for the high cost. We propose a model for the occurrence of such subproblems when using the backtracking algorithm. We calculate the probability of their occurrence and estimate their cost. From this, we derive the theoretical cost distribution when the constraint graph is complete and show that it matches the observed cost distribution in this region. We suggest that a similar model would also account for the exceptionally hard problems that have been observed using more sophisticated algorithms.
Unable to display preview. Download preview PDF.
- 1.A. B. Baker. Intelligent Backtracking on the Hardest Constraint Problems. Technical report, CIRL, University of Oregon, Feb. 1995.Google Scholar
- 2.R. J. Bayardo and R. Schrag. Using CSP Look-Back Techniques to Solve Exceptionally Hard SAT Instances. In FProceedings CP96, pages 46–60, 1996.Google Scholar
- 3.P. Cheeseman, B. Kanefsky, and W. Taylor.Where the Really Hard Problems are. In Proceedings IJCAI-91, volume 1, pages 331–337, 1991.Google Scholar
- 4.I. P. Gent and T. Walsh. Easy Problems are Sometimes Hard. Artificial Intelligence, 70:335–345, 1994.Google Scholar
- 5.I. P. Gent and T. Walsh. Phase Transitions from Real Computational Problems. In Proceedings of the 8t1î International Symposium on Artificial Intelligence, pages 356–364, Oct. 1995.Google Scholar
- 6.R. Haralick and G. Elliott. Increasing tree search efficiency for constraint satisfaction problems. Artificial Intelligence, 14:263–313, 1980.Google Scholar
- 7.T. Hogg and C. P. Williams. The Hardest Constraint Problems: A Double Phase Transition. Artificial Intelligence, 69:359–377, 1994.Google Scholar
- 8.P. Prosser. Hybrid Algorithms for the Constraint Satisfaction Problem. Computational Intelligence, 9(3):268–299, 1993.Google Scholar
- 9.B. Selman and S. Kirkpatrick. Critical behaviour in the computational cost of satisfiability testing. Artificial Intelligence, 81:273–295, 1996.Google Scholar
- 10.B. M. Smith and S. A. Grant. Sparse Constraint Graphs and Exceptionally Hard Problems. In Proceedings IJCAI95, volume 1, pages 646–651, 1995.Google Scholar