Restart Strategy Selection Using Machine Learning Techniques

Authors:
Shai Haim;Toby Walsh
Affiliations:
NICTA and School of Computer Science and Engineering, University of New South Wales, Sydney, Australia;NICTA and School of Computer Science and Engineering, University of New South Wales, Sydney, Australia
Venue:
SAT '09 Proceedings of the 12th International Conference on Theory and Applications of Satisfiability Testing
Year:
2009

Citing 18
Cited 1

Boosting combinatorial search through randomization

AAAI '98/IAAI '98 Proceedings of the fifteenth national/tenth conference on Artificial intelligence/Innovative applications of artificial intelligence
Chaff: engineering an efficient SAT solver

Proceedings of the 38th annual Design Automation Conference
Restart Policies with Dependence among Runs: A Dynamic Programming Approach

CP '02 Proceedings of the 8th International Conference on Principles and Practice of Constraint Programming
Dynamic restart policies

Eighteenth national conference on Artificial intelligence
BerkMin: A Fast and Robust Sat-Solver

Proceedings of the conference on Design, automation and test in Europe
Sparse Multinomial Logistic Regression: Fast Algorithms and Generalization Bounds

IEEE Transactions on Pattern Analysis and Machine Intelligence
SATzilla: portfolio-based algorithm selection for SAT

Journal of Artificial Intelligence Research
Search in a small world

IJCAI'99 Proceedings of the 16th international joint conference on Artificial intelligence - Volume 2
The effect of restarts on the efficiency of clause learning

IJCAI'07 Proceedings of the 20th international joint conference on Artifical intelligence
On universal restart strategies for backtracking search

CP'07 Proceedings of the 13th international conference on Principles and practice of constraint programming
Hierarchical hardness models for SAT

CP'07 Proceedings of the 13th international conference on Principles and practice of constraint programming
A case for simple SAT solvers

CP'07 Proceedings of the 13th international conference on Principles and practice of constraint programming
Adaptive restart strategies for conflict driven SAT solvers

SAT'08 Proceedings of the 11th international conference on Theory and applications of satisfiability testing
Attacking Bivium using SAT solvers

SAT'08 Proceedings of the 11th international conference on Theory and applications of satisfiability testing
Online estimation of SAT solving runtime

SAT'08 Proceedings of the 11th international conference on Theory and applications of satisfiability testing
Local restarts

SAT'08 Proceedings of the 11th international conference on Theory and applications of satisfiability testing
Summarizing CSP hardness with continuous probability distributions

AAAI'97/IAAI'97 Proceedings of the fourteenth national conference on artificial intelligence and ninth conference on Innovative applications of artificial intelligence
Performance prediction and automated tuning of randomized and parametric algorithms

CP'06 Proceedings of the 12th international conference on Principles and Practice of Constraint Programming

Assisting in search heuristics selection through multidimensional supervised classification: A case study on software testing

Information Sciences: an International Journal

Quantified Score

Hi-index	0.00

Visualization

Abstract

Restart strategies are an important factor in the performance of conflict-driven Davis Putnam style SAT solvers. Selecting a good restart strategy for a problem instance can enhance the performance of a solver. Inspired by recent success applying machine learning techniques to predict the runtime of SAT solvers, we present a method which uses machine learning to boost solver performance through a smart selection of the restart strategy. Based on easy to compute features, we train both a satisfiability classifier and runtime models. We use these models to choose between restart strategies. We present experimental results comparing this technique with the most commonly used restart strategies. Our results demonstrate that machine learning is effective in improving solver performance.