Automatic construction of parallel portfolios via algorithm configuration

作者：Lindauer Marius^*; Hoos Holger^*; Leyton Brown Kevin^*; Schaub Torsten^*

来源：Artificial Intelligence, 2017, 244: 272-290.

DOI：10.1016/j.artint.2016.05.004

• Algorithm configuration
• Parallel SAT solving
• Algorithm portfolios
• Programming by optimization
• Automated parallelization

Since 2004, increases in computational power described by Moore's law have substantially been realized in the form of additional cores rather than through faster clock speeds. To make effective use of modern hardware when solving hard computational problems, it is therefore necessary to employ parallel solution strategies. In this work, we demonstrate how effective parallel solvers for propositional satisfiability (SAT), one of the most widely studied NP-complete problems, can be produced automatically from any existing sequential, highly parametric SAT solver. Our Automatic Construction of Parallel Portfolios (ACPP) approach uses an automatic algorithm configuration procedure to identify a set of configurations that perform well when executed in parallel. Applied to two prominent SAT solvers, Lingeling and clasp, our ACPP procedure identified 8-core solvers that significantly outperformed their sequential counterparts on a diverse set of instances from the application and hard combinatorial category of the 2012 SAT Challenge. We further extended our ACPP approach to produce parallel portfolio solvers consisting of several different solvers by combining their configuration spaces. Applied to the component solvers of the 2012 SAT Challenge gold medal winning SAT Solver pfolioUZK, our ACPP procedures produced a significantly better-performing parallel SAT solver.

• 出版日期2017-3