Using the unconstrained quadratic program to model and solve Max 2-SAT problems
نویسندگان
چکیده
Satisfiability (SAT) and Max-SAT problems have been the object of considerable research effort over the past few decades. They remain a very important research area today due to their computational challenge and application importance. In this paper we investigate the use of penalty functions to recast SAT problems into the modelling framework offered by the unconstrained quadratic binary program. Computational experience is presented, illustrating how promising this approach is for Max 2-Sat problems.
منابع مشابه
A Penalized Quadratic Convex Reformulation Method for Random Quadratic Unconstrained Binary Optimization
The Quadratic Convex Reformulation (QCR) method is used to solve quadratic unconstrained binary optimization problems. In this method, the semidefinite relaxation is used to reformulate it to a convex binary quadratic program which is solved using mixed integer quadratic programming solvers. We extend this method to random quadratic unconstrained binary optimization problems. We develop a Penal...
متن کاملSolving MAX-SAT and Weighted MAX-SAT Problems Using Branch-and-Cut
We describe a branch and cut algorithm for both MAX-SAT and weighted MAX-SAT. This algorithm uses the GSAT procedure as a primal heuristic. At each node we solve a linear programming (LP) relaxation of the problem. Two styles of separating cuts are added: resolution cuts and odd cycle inequalities. We compare our algorithm to an extension of the Davis Putnam Loveland (EDPL) algorithm and a Semi...
متن کاملA branch and cut algorithm for MAX-SAT and weighted MAX-SAT
We describe a branch and cut algorithm for both MAX-SAT and weighted MAX-SAT. This algorithm uses the GSAT procedure as a primal heuristic. At each node we solve a linear programming (LP) relaxation of the problem. Two styles of separating cuts are added: resolution cuts and odd cycle inequalities. We compare our algorithm to an extension of the Davis Putnam Loveland (EDPL) algorithm. Our algor...
متن کاملComparative Studies of Constraint
Maximum satissability (MAX-SAT) is an extension of satissability (SAT), in which a partial solution is sought that satisses the maximum number of clauses in a logical formula. Enumerative methods giving guaranteed optimal solutions can be derived from traditional search algorithms used to solve SAT problems, in particular the Davis-Putnam procedure. Algorithms have also been developed for the m...
متن کاملA Comparison of Methods for Solving MAX-SAT Problems
We compare the performance of 3 approaches to the solution of MAX-SAT problems, including a version of the DavisPutnam-Loveland algorithm extended to solve MAX-SAT, an integer programming branch-and-cut algorithm and an algorithm for MAX-2-SAT problems based on a semide nite programming relation.
متن کامل