English

On the Mysteries of MAX NAE-SAT

Computational Complexity 2024-09-27 v2 Data Structures and Algorithms

Abstract

MAX NAE-SAT is a natural optimization problem, closely related to its better-known relative MAX SAT. The approximability status of MAX NAE-SAT is almost completely understood if all clauses have the same size kk, for some k2k\ge 2. We refer to this problem as MAX NAE-{k}\{k\}-SAT. For k=2k=2, it is essentially the celebrated MAX CUT problem. For k=3k=3, it is related to the MAX CUT problem in graphs that can be fractionally covered by triangles. For k4k\ge 4, it is known that an approximation ratio of 112k11-\frac{1}{2^{k-1}}, obtained by choosing a random assignment, is optimal, assuming PNPP\ne NP. For every k2k\ge 2, an approximation ratio of at least 78\frac{7}{8} can be obtained for MAX NAE-{k}\{k\}-SAT. There was some hope, therefore, that there is also a 78\frac{7}{8}-approximation algorithm for MAX NAE-SAT, where clauses of all sizes are allowed simultaneously. Our main result is that there is no 78\frac{7}{8}-approximation algorithm for MAX NAE-SAT, assuming the unique games conjecture (UGC). In fact, even for almost satisfiable instances of MAX NAE-{3,5}\{3,5\}-SAT (i.e., MAX NAE-SAT where all clauses have size 33 or 55), the best approximation ratio that can be achieved, assuming UGC, is at most 3(214)20.8739\frac{3(\sqrt{21}-4)}{2}\approx 0.8739. Using calculus of variations, we extend the analysis of O'Donnell and Wu for MAX CUT to MAX NAE-{3}\{3\}-SAT. We obtain an optimal algorithm, assuming UGC, for MAX NAE-{3}\{3\}-SAT, slightly improving on previous algorithms. The approximation ratio of the new algorithm is 0.9089\approx 0.9089. We complement our theoretical results with some experimental results. We describe an approximation algorithm for almost satisfiable instances of MAX NAE-{3,5}\{3,5\}-SAT with a conjectured approximation ratio of 0.8728, and an approximation algorithm for almost satisfiable instances of MAX NAE-SAT with a conjectured approximation ratio of 0.8698.

Keywords

Cite

@article{arxiv.2009.10677,
  title  = {On the Mysteries of MAX NAE-SAT},
  author = {Joshua Brakensiek and Neng Huang and Aaron Potechin and Uri Zwick},
  journal= {arXiv preprint arXiv:2009.10677},
  year   = {2024}
}

Comments

44 pages, 8 figures, accepted to SIDMA