Stochastic Local Search Methods for Dynamic SAT- an Initial Investigation

作者: Holger H. Hoos , Kevin O''Neill

DOI:

关键词: SimplicityPropositional calculusBoolean satisfiability problemDynamic problemMathematical optimizationMathematicsSat problemLocal search (optimization)WalkSAT

摘要: We introduce the dynamic SAT problem, a generalisation of satisfiability problem in propositional logic which allows changes over time. DynSAT can be seen as particular form CSP, but considering results and recent success solving conventional problems, we believe that conceptual simplicity will allow us to more easily devise investigate high-performing algorithms for than CSPs. In this article, motivate discuss various formalisations it, stochastic local search (SLS) it. particular, apply SLS perform well on problems analyse characterise their performance empirically; initial investigation indicates differences between well-known WalkSAT family generally carry when applied DynSAT. also study different generic approaches using types problems.

参考文章(14)
Rina Dechter, Avi Dechter, Belief maintenance in dynamic constraint networks national conference on artificial intelligence. pp. 37- 42 ,(1988)
Holger H. Hoos, On the run-time behaviour of stochastic local search algorithms for SAT national conference on artificial intelligence. pp. 661- 666 ,(1999)
Richard J. Wallace, Eugene C. Freuder, Stable Solutions for Dynamic Constraint Satisfaction Problems principles and practice of constraint programming. pp. 447- 461 ,(1998) , 10.1007/3-540-49481-2_32
Bart Selman, Henry Kautz, Pushing the envelope: planning, propositional logic, and stochastic search national conference on artificial intelligence. pp. 1194- 1201 ,(1996)
Bart Selman, David Mitchell, Hector Levesque, A new method for solving hard satisfiability problems national conference on artificial intelligence. pp. 440- 446 ,(1992)
Holger H. Hoos, Thomas Stützle, Towards a characterisation of the behaviour of stochastic local search algorithms for SAT Artificial Intelligence. ,vol. 112, pp. 213- 232 ,(1999) , 10.1016/S0004-3702(99)00048-X
Bart Selman, Henry Kautz, David McAllester, Evidence for invariants in local search national conference on artificial intelligence. pp. 321- 326 ,(1997)
Steven Minton, Mark D. Johnston, Andrew B. Philips, Philip Laird, Minimizing conflicts: a heuristic repair method for constraint satisfaction and scheduling problems Artificial Intelligence. ,vol. 58, pp. 161- 205 ,(1992) , 10.1016/0004-3702(92)90007-K
Andrew J. Parkes, Amitabha Roy, Matthew L. Ginsberg, Supermodels and robustness national conference on artificial intelligence. pp. 334- 339 ,(1998)
Holger H. Hoos, Thomas Stützle, Evaluating las vegas algorithms: pitfalls and remedies uncertainty in artificial intelligence. pp. 238- 245 ,(1998)