Max-2-SAT

Max-2-SAT
Name	Max-2-SAT
Field	Theoretical computer science, Combinatorial optimization
Problem	Boolean satisfiability variant
Input	2-CNF formula, integer k
Goal	maximize number of satisfied clauses
Complexity	NP-hard, APX-complete

Max-2-SAT is an optimization variant of Boolean satisfiability focusing on formulas in 2-conjunctive normal form. It asks, given a collection of clauses each containing at most two literals, to find an assignment that satisfies the maximum number of clauses, and it sits at the intersection of computational complexity, approximation theory, and combinatorial optimization.

Definition and problem statement

Max-2-SAT is defined for a finite set of Boolean variables and a multiset of clauses each with one or two literals drawn from those variables. Instances are often presented as 2-CNF formulas with clause weights and the objective is to maximize the sum of weights of satisfied clauses under a truth assignment. Standard decision and optimization formulations include asking whether at least k clauses can be satisfied, or computing an assignment achieving the optimal value; these formulations are used in reductions, proofs, and algorithm design in texts and venues associated with Stephen Cook, Richard Karp, Michael Garey, David Johnson, and research groups at institutions such as MIT, Stanford University, and Princeton University.

Complexity and computational hardness

Max-2-SAT is NP-hard and remains NP-hard under restrictions such as planar incidence graphs or bounded variable occurrences; hardness proofs draw on reductions associated with the classical NP-complete problems catalogued by Karp's 21 NP-complete problems and developments in complexity theory by Cook–Levin theorem influences and later refinements by Arora and Schoening. The decision variant ("at least k clauses") is NP-complete, and approximation hardness results place Max-2-SAT in APX-complete or Unique Games-related frameworks studied by researchers linked to Subhash Khot, Sanjeev Arora, Madhu Sudan, Irit Dinur, and the PCP theorem literature. Inapproximability thresholds have been tightened using probabilistically checkable proof techniques associated with work at Princeton University and Microsoft Research.

Approximation algorithms and bounds

Approximation algorithms for Max-2-SAT include randomized rounding and semidefinite programming methods inspired by breakthroughs such as the Goemans–Williamson algorithm and subsequent refinements by groups at Bell Labs, Bellcore, and academic labs including University of California, Berkeley. A simple randomized assignment achieves expected approximation ratios demonstrable in analyses by Michael Luby and Noga Alon, while deterministic and derandomized approaches exploit method-of-pessimistic estimators or conditional expectations credited to researchers at IBM Research and AT&T Laboratories. Semidefinite programming relaxations yield improved ratios following techniques attributed to Prasad Raghavendra, Umesh Vazirani, and Michel Goemans, and tight bounds relate to integrality gaps studied by scholars at ETH Zurich and Carnegie Mellon University.

Exact algorithms and parameterized approaches

Exact exponential-time algorithms for Max-2-SAT exploit branching, inclusion–exclusion, and measure-and-conquer analyses developed in the algorithmic communities around Donald Knuth, Richard Karp, and research groups at Tel Aviv University and University of Warsaw. Parameterized complexity approaches treat parameters such as the number of unsatisfied clauses, solution value above guarantee, or structural measures (treewidth, pathwidth) with fixed-parameter tractable algorithms influenced by work from Rod Downey, Michael Fellows, Jianer Chen, and teams at Chinese Academy of Sciences and University of Helsinki. Kernelization and bounded-search-tree techniques reduce instance sizes and run in times analyzed in conferences like STOC, FOCS, and SODA.

Special cases and variants

Variants include weighted Max-2-SAT, unweighted Max-2-SAT, Max-Cut equivalences under variable transformations studied by researchers linked to William Kinnersley and classical graph algorithmists at Cornell University. Special cases on restricted graph classes—planar, bipartite, bounded-degree, or bounded-clique-width—admit different complexity and approximation behaviours analyzed by scholars at University of Cambridge and University of Oxford. Related problems and generalizations such as Max-k-SAT, Min-2-SAT, and constraint satisfaction problems connect to theoretical frameworks by Thomas Feder, Moshe Y. Vardi, and the complexity landscape described in surveys from DIMACS workshops and summer schools at Institute for Advanced Study.

Applications and practical implementations

Max-2-SAT appears in practical contexts including hardware verification and FPGA mapping studied by teams at Xerox PARC, Intel, and IBM, in bioinformatics and protein interaction modeling with contributions from labs at Broad Institute and European Bioinformatics Institute, and in statistical inference tasks that intersect with machine learning research at Google Research and DeepMind. Implementations leverage SAT solvers, local search heuristics, and semidefinite programming libraries originating from groups at Google, Microsoft Research Cambridge, and open-source projects coordinated via repositories and communities associated with GitHub and international algorithm competitions such as the DIMACS Implementation Challenge.

Category:Computational complexity