Average-case analysis for the MAX-2SAT problem

Authors:
Osamu Watanabe;Masaki Yamamoto
Affiliations:
Department of Math. and Comp. Sci., Tokyo Inst. of Technology, Japan;Department of Mathematical Sciences, School of Science, Tokai University, Japan
Venue:
Theoretical Computer Science
Year:
2010

Citing 9
Cited 0

Probabilistic reasoning in intelligent systems: networks of plausible inference

Probabilistic reasoning in intelligent systems: networks of plausible inference
Randomized algorithms

Randomized algorithms
Some optimal inapproximability results

Journal of the ACM (JACM)
Scatter Search with Random Walk Strategy for SAT and MAX-W-SAT Problems

Proceedings of the 14th International conference on Industrial and engineering applications of artificial intelligence and expert systems: engineering of intelligent systems
Generating Instances for MAX2SAT with Optimal Solutions

Theory of Computing Systems
Hiding satisfying assignments: two are better than one

Journal of Artificial Intelligence Research
Tabu search for SAT

AAAI'97/IAAI'97 Proceedings of the fourteenth national conference on artificial intelligence and ninth conference on Innovative applications of artificial intelligence
Efficient and experimental meta-heuristics for MAX-SAT problems

WEA'05 Proceedings of the 4th international conference on Experimental and Efficient Algorithms
Turbo decoding as an instance of Pearl's “belief propagation” algorithm

IEEE Journal on Selected Areas in Communications

Quantified Score

Hi-index	5.23

Visualization

Abstract

We propose a simple probability model for MAX-2SAT instances for discussing the average-case complexity of the MAX-2SAT problem. Our model is a ''planted solution model'', where each instance is generated randomly from a target solution. We show that for a large range of parameters, the planted solution (more precisely, one of the planted solution pairs) is the optimal solution for the generated instance with high probability. We then give a simple linear-time algorithm based on a message passing method, and we prove that it solves the MAX-2SAT problem with high probability for random MAX-2SAT instances under this planted solution model for probability parameters within a certain range.