TY - GEN
T1 - A Novel Algorithm for Max Sat Calling MOCE to Order
AU - Berend, Daniel
AU - Golan, Shahar
AU - Twitto, Yochai
N1 - Publisher Copyright:
© 2021, Springer Nature Switzerland AG.
PY - 2021/1/1
Y1 - 2021/1/1
N2 - In this paper, we present and study a new algorithm for the Maximum Satisfiability (Max Sat) problem. The algorithm, GO-MOCE, is based on the Method of Conditional Expectations (MOCE, also known as Johnson’s Algorithm), and applies a greedy variable ordering to it. We conduct an extensive empirical evaluation on two collections of instances – instances from a past Max Sat competition and random instances. We show that GO-MOCE reduces the number of unsatisfied clauses by tens of percents as compared to MOCE. We prove that, using tailored data structures we designed, GO-MOCE retains the linear time complexity. Moreover, its runtime overhead in our experiments is at most 10%. We combine GO-MOCE with CCLS, a state-of-the-art solver, and show that the combined solver improves CCLS on the above mentioned collections.
AB - In this paper, we present and study a new algorithm for the Maximum Satisfiability (Max Sat) problem. The algorithm, GO-MOCE, is based on the Method of Conditional Expectations (MOCE, also known as Johnson’s Algorithm), and applies a greedy variable ordering to it. We conduct an extensive empirical evaluation on two collections of instances – instances from a past Max Sat competition and random instances. We show that GO-MOCE reduces the number of unsatisfied clauses by tens of percents as compared to MOCE. We prove that, using tailored data structures we designed, GO-MOCE retains the linear time complexity. Moreover, its runtime overhead in our experiments is at most 10%. We combine GO-MOCE with CCLS, a state-of-the-art solver, and show that the combined solver improves CCLS on the above mentioned collections.
UR - http://www.scopus.com/inward/record.url?scp=85121838951&partnerID=8YFLogxK
U2 - 10.1007/978-3-030-92681-6_25
DO - 10.1007/978-3-030-92681-6_25
M3 - Conference contribution
AN - SCOPUS:85121838951
SN - 9783030926809
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 302
EP - 317
BT - Combinatorial Optimization and Applications - 15th International Conference, COCOA 2021, Proceedings
A2 - Du, Ding-Zhu
A2 - Du, Donglei
A2 - Wu, Chenchen
A2 - Xu, Dachuan
PB - Springer Science and Business Media Deutschland GmbH
T2 - 15th Annual International Conference on Combinatorial Optimization and Applications, COCOA 2021
Y2 - 17 December 2021 through 19 December 2021
ER -