Symbiosis of Search and Heuristics for Random 3-SAT

Symbiosis of Search and Heuristics for Random 3-SAT
复制标题

随机 3-SAT 的搜索和启发式共生

DOI:
--
复制
发表时间:
2014
期刊:
arXiv.org
影响因子:
--
通讯作者:
Marijn J. H. Heule
Marijn J. H. Heule
中科院分区:
--
文献类型:
--
作者:
S. Mijnders;B. D. Wilde;Marijn J. H. Heule

文献摘要

被引文献

相似文献

代尔夫特理工大学,代尔夫特,荷兰当正确组合时,搜索技术可以揭示复杂的分支启发法的全部潜力。我们在著名的随机 3-SAT 公式中证明了这一观察结果。首先,提出了一种新的分支启发式,它概括了此类的现有工作。使用这种启发式可以构建更小的搜索树。其次,我们介绍差异搜索的一种变体,称为 ALDS。理论和实践证据表明,当与新的启发式方法相结合时,ALDS 会以接近最优的顺序遍历搜索树。搜索和启发式这两种技术都已在前瞻求解器中实现。 SAT 2009 竞赛结果表明,march 是迄今为止随机 k-SAT 公式上最强的完整求解器。
Delft University of Technology, Delft, The NetherlandsAbstract. When combined properly, search techniques can reveal the full poten-tial of sophisticated branching heuristics. We demonstrate this observation on thewell-known class of random 3-SAT formulae. First, a new branching heuristic ispresented, which generalizes existing work on this class. Much smaller searchtrees can be constructed by using this heuristic. Second, we introduce a variantof discrepancy search, called ALDS. Theoretical and practical evidence supportthat ALDS traverses the search tree in a near-optimal order when combined withthe new heuristic. Both techniques, search and heuristic, have been implementedin the look-ahead solvermarch. The SAT 2009 competition results show thatmarch is by far the strongest complete solver on random k-SAT formulae.