An SVM-Based Prediction Method for Solving SAT Problems

We show how Support vector machines (SVM) can be applied to the Satisfiability (SAT) problem and how their prediction results can be naturally applied to both incomplete and complete SAT solvers. SVM is used for the classification of the variables in the SAT problem and the classification results are the assignment of the variables. And we also present empirical results of applying SVM to instances of the SAT problem from the Center for Discrete Mathematics and Theoretical Computer Science (DIMACS) archive and compare them against the results of other incomplete and complete algorithms for the SAT problem.