Boosting SLS Performance by Incorporating Resolution-based Preprocessor
نویسندگان
چکیده
State of the art Stochastic Local Search (SLS) solvers have difficulty in solving many CNF-encoded realistic SAT problems, apparently because they are unable to exploit hidden structure as well as systematic solvers. Recent work has shown that SLS solvers may benefit from a preprocessing phase borrowed from systematic SAT solving. In this paper, we report an extensive empirical examination of the impact of SAT preprocessing on the performance of contemporary SLS solvers. It emerges that all the examined solvers do indeed benefit from preprocessing, and the effect of each preprocessor is close to uniform across solvers and across problems. Our results suggest that SLS solvers need to be equipped with multiple preprocessors if they are ever to match the performance of systematic solvers on highly structured problems.
منابع مشابه
Boosting the Performance of SLS and CDCL Solvers by Preprocessor Tuning
Preprocessing techniques are crucial for SAT solvers when it comes to reaching state-of-the-art performance as it was shown by the results of the last SAT Competitions. The usefulness of a preprocessing technique depends highly on its own parameters, on the instances on which it is applied and on the used solver. In this paper we first give an extended analysis of the performance gain reached b...
متن کاملResolution Enhanced SLS solvers: R+PAWS, R+RSAPS and R+ANOV+
Recent work on Stochastic Local Search (SLS) for the SAT and CSP domains has shown the superior performance of SLS over traditional backtracking algorithms on a broad range of problem instances. In this paper, we report on a technique for enhancing the performance of SLS solvers by incorporating a preprocessing phase in which resolution is used to deduce consequences of the input clauses, expos...
متن کاملStructured least squares to improve the performance of ESPRIT-type high-resolution techniques
Absfraet ESPRIT-type high-resolution (spatial) frequency estimation techniques, like standard ESPRIT, state space methods, matrix pencil methods, or Unitary ESPRIT, obtain their (spatial) frequency estimates from the solution of a highly-structured, overdetermined system of equations. Here, the structure is defined in terms of two selection matrices applied to a matrix spanning the estimated si...
متن کاملQuality Adaptive Low-Rank Based JPEG Decoding with Applications
Small compression noises, despite being transparent to human eyes, can adversely affect the results of many image restoration processes, if left unaccounted for. Especially, compression noises are highly detrimental to inverse operators of high-boosting (sharpening) nature, such as deblurring and superresolution against a convolution kernel. By incorporating the non-linear DCT quantization mech...
متن کاملIncorporating Prior Knowledge into Boosting for Multi-Label Classification XiaoWang
Multi-label learning deals with the problem where each instance may belong to multiple labels simultaneously. The task of the learning paradigm is to output the label set whose size is unknown a priori for each unseen instance, through analyzing the training data set with known label sets. Existing multi-label learning algorithms are almost based on the purely data-driven method. The larger the...
متن کامل