Refuting conjectures in extremal combinatorics via linear programming

Adam Zsolt Wagner

DOI: https://doi.org/10.1016/j.jcta.2019.105130

2020-01-01

Abstract:<p>We apply simple linear programming methods and an LP solver to refute a number of open conjectures in extremal combinatorics.</p>

What problem does this paper attempt to address?

On modeling hard combinatorial optimization problems as linear programs: Refutations of the "unconditional impossibility" claims

Moustapha Diaby,Mark H. Karwan,Lei Sun

DOI: https://doi.org/10.48550/arXiv.1902.03549

2019-02-10

Computational Complexity

Abstract:There has been a series of developments in the recent literature (by essentially a same "circle" of authors) with the absolute/unconditioned (implicit or explicit) claim that there exists no abstraction of an NP-Complete combinatorial optimization problem in which the defining combinatorial configurations (such as "tours" in the case of the traveling salesman problem (TSP) for example) can be modeled by a polynomial-sized system of linear constraints. The purpose of this paper is to provide general as well as specific refutations for these recent claims.
Linear Extension of the Erdős-Heilbronn Conjecture.

Zhi-Wei Sun,Li-Lu Zhao

DOI: https://doi.org/10.1016/j.jcta.2011.09.003

IF: 1.263

2008-01-01

Journal of Combinatorial Theory Series A

Abstract:The famous Erdős–Heilbronn conjecture plays an important role in the development of additive combinatorial number theory. In 2007 Z.W. Sun made the following further conjecture (which is the linear extension of the Erdős–Heilbronn conjecture): For any finite subset A of a field F and nonzero elements a1,…,an of F, we have|{a1x1+⋯+anxn:x1,…,xn∈A,andxi≠xjifi≠j}|⩾min{p(F)−δ,n(|A|−n)+1}, where the additive order p(F) of the multiplicative identity of F is different from n+1, and δ∈{0,1} takes the value 1 if and only if n=2 and a1+a2=0. In this paper we prove this conjecture of Sun when p(F)⩾n(3n−5)/2. We also obtain a sharp lower bound for the cardinality of the restricted sumset{x1+⋯+xn:x1∈A1,…,xn∈An,andP(x1,…,xn)≠0}, where A1,…,An are finite subsets of a field F and P(x1,…,xn) is a general polynomial over F.
A Combinatorial Certifying Algorithm for Linear Programming Problems with Gainfree Leontief Substitution Systems

Kei Kimura,Kazuhisa Makino

2023-06-06

Abstract:Linear programming (LP) problems with gainfree Leontief substitution systems have been intensively studied in economics and operations research, and include the feasibility problem of a class of Horn systems, which arises in, e.g., polyhedral combinatorics and logic. This subclass of LP problems admits a strongly polynomial time algorithm, where devising such an algorithm for general LP problems is one of the major theoretical open questions in mathematical optimization and computer science. Recently, much attention has been paid to devising certifying algorithms in software engineering, since those algorithms enable one to confirm the correctness of outputs of programs with simple computations. In this paper, we provide the first combinatorial (and strongly polynomial time) certifying algorithm for LP problems with gainfree Leontief substitution systems. As a by-product, we answer affirmatively an open question whether the feasibility problem of the class of Horn systems admits a combinatorial certifying algorithm.

Data Structures and Algorithms
Some Zero-One Linear Programming Reformulations for the Maximum Clique Problem

Ákos Beke,Sándor Szabó,Bogdán Zavalnij

DOI: https://doi.org/10.1556/314.2020.00005

2021-04-08

Mathematica Pannonica

Abstract:Many combinatorial optimization problems can be expressed in terms of zero-one linear programs. For the maximum clique problem the so-called edge reformulation is applied most commonly. Two less frequently used LP equivalents are the independent set and edge covering set reformulations. The number of the constraints (as a function of the number of vertices of the ground graph) is asymptotically quadratic in the edge and the edge covering set LP reformulations and it is exponential in the independent set reformulation, respectively. F. D. Croce and R. Tadei proposed an approach in which the number of the constraints is equal to the number of the vertices. In this paper we are looking for possible tighter variants of these linear programs.
Sparsified Linear Programming for Zero-Sum Equilibrium Finding

Brian Hu Zhang,Tuomas Sandholm

DOI: https://doi.org/10.48550/arXiv.2006.03451

2020-06-30

Abstract:Computational equilibrium finding in large zero-sum extensive-form imperfect-information games has led to significant recent AI breakthroughs. The fastest algorithms for the problem are new forms of counterfactual regret minimization [Brown and Sandholm, 2019]. In this paper we present a totally different approach to the problem, which is competitive and often orders of magnitude better than the prior state of the art. The equilibrium-finding problem can be formulated as a linear program (LP) [Koller et al., 1994], but solving it as an LP has not been scalable due to the memory requirements of LP solvers, which can often be quadratically worse than CFR-based algorithms. We give an efficient practical algorithm that factors a large payoff matrix into a product of two matrices that are typically dramatically sparser. This allows us to express the equilibrium-finding problem as a linear program with size only a logarithmic factor worse than CFR, and thus allows linear program solvers to run on such games. With experiments on poker endgames, we demonstrate in practice, for the first time, that modern linear program solvers are competitive against even game-specific modern variants of CFR in solving large extensive-form games, and can be used to compute exact solutions unlike iterative algorithms like CFR.

Computer Science and Game Theory,Machine Learning
Three conjectures in extremal spectral graph theory

Michael Tait,Josh Tobin

DOI: https://doi.org/10.48550/arXiv.1606.01916

2017-04-22

Abstract:We prove three conjectures regarding the maximization of spectral invariants over certain families of graphs. Our most difficult result is that the join of $P_2$ and $P_{n-2}$ is the unique graph of maximum spectral radius over all planar graphs. This was conjectured by Boots and Royle in 1991 and independently by Cao and Vince in 1993. Similarly, we prove a conjecture of Cvetković and Rowlinson from 1990 stating that the unique outerplanar graph of maximum spectral radius is the join of a vertex and $P_{n-1}$. Finally, we prove a conjecture of Aouchiche et al from 2008 stating that a pineapple graph is the unique connected graph maximizing the spectral radius minus the average degree. To prove our theorems, we use the leading eigenvector of a purported extremal graph to deduce structural properties about that graph. Using this setup, we give short proofs of several old results: Mantel's Theorem, Stanley's edge bound and extensions, the Kővari-Sós-Turán Theorem applied to $\mathrm{ex}\left(n, K_{2,t}\right)$, and a partial solution to an old problem of Erdős on making a triangle-free graph bipartite.

Combinatorics
Zero-Sum Games and Linear Programming Duality

Bernhard von Stengel

DOI: https://doi.org/10.1287/moor.2022.0149

2023-10-22

Abstract:The minimax theorem for zero-sum games is easily proved from the strong duality theorem of linear programming. For the converse direction, the standard proof by Dantzig (1951) is known to be incomplete. We explain and combine classical theorems about solving linear equations with nonnegative variables to give a correct alternative proof, more directly than Adler (2013). We also extend Dantzig's game so that any max-min strategy gives either an optimal LP solution or shows that none exists.

Computer Science and Game Theory,Optimization and Control
On the Power of Symmetric Linear Programs

Albert Atserias,Anuj Dawar,Joanna Ochremiak

DOI: https://doi.org/10.48550/arXiv.1901.07825

2019-01-23

Abstract:We consider families of symmetric linear programs (LPs) that decide a property of graphs (or other relational structures) in the sense that, for each size of graph, there is an LP defining a polyhedral lift that separates the integer points corresponding to graphs with the property from those corresponding to graphs without the property. We show that this is equivalent, with at most polynomial blow-up in size, to families of symmetric Boolean circuits with threshold gates. In particular, when we consider polynomial-size LPs, the model is equivalent to definability in a non-uniform version of fixed-point logic with counting (FPC). Known upper and lower bounds for FPC apply to the non-uniform version. In particular, this implies that the class of graphs with perfect matchings has polynomial-size symmetric LPs while we obtain an exponential lower bound for symmetric LPs for the class of Hamiltonian graphs. We compare and contrast this with previous results (Yannakakis 1991) showing that any symmetric LPs for the matching and TSP polytopes have exponential size. As an application, we establish that for random, uniformly distributed graphs, polynomial-size symmetric LPs are as powerful as general Boolean circuits. We illustrate the effect of this on the well-studied planted-clique problem.

Logic in Computer Science,Computational Complexity,Optimization and Control
0-1 Integer Linear Programming with a Linear Number of Constraints

Russell Impagliazzo,Shachar Lovett,Ramamohan Paturi,Stefan Schneider

DOI: https://doi.org/10.48550/arXiv.1401.5512

2014-01-21

Computational Complexity

Abstract:We give an exact algorithm for the 0-1 Integer Linear Programming problem with a linear number of constraints that improves over exhaustive search by an exponential factor. Specifically, our algorithm runs in time $2^{(1-\text{poly}(1/c))n}$ where n is the number of variables and cn is the number of constraints. The key idea for the algorithm is a reduction to the Vector Domination problem and a new algorithm for that subproblem.
Conic programming to understand sums of squares of eigenvalues of graphs

Gabriel Coutinho,Thomás Jung Spier,Shengtong Zhang

2024-11-13

Abstract:In this paper we prove a conjecture by Wocjan, Elphick and Anekstein (2018) which upper bounds the sum of the squares of the positive (or negative) eigenvalues of the adjacency matrix of a graph by an expression that behaves monotonically in terms of the vector chromatic number. One of our lemmas is a strengthening of the Cauchy-Schwarz inequality for Hermitian matrices when one of the matrices is positive semidefinite. A related conjecture due to Bollobás and Nikiforov (2007) replaces the vector chromatic number by the clique number and sums over the first two eigenvalues only. We prove a version of this conjecture with weaker constants. An important consequence of our work is a proof that for any fixed $r$, computing a rank $r$ optimum solution to the vector chromatic number semidefinite programming is NP-hard. We also present a vertex weighted version of some of our results, and we show how it leads quite naturally to the known vertex-weighted version of the Motzkin-Straus quadratic optimization formulation for the clique number.

Combinatorics,Optimization and Control,Spectral Theory
Analysis of the Min-Sum Algorithm for Packing and Covering Problems via Linear Programming

Guy Even,Nissim Halabi

DOI: https://doi.org/10.48550/arXiv.1302.3518

2013-07-15

Abstract:Message-passing algorithms based on belief-propagation (BP) are successfully used in many applications including decoding error correcting codes and solving constraint satisfaction and inference problems. BP-based algorithms operate over graph representations, called factor graphs, that are used to model the input. Although in many cases BP-based algorithms exhibit impressive empirical results, not much has been proved when the factor graphs have cycles. This work deals with packing and covering integer programs in which the constraint matrix is zero-one, the constraint vector is integral, and the variables are subject to box constraints. We study the performance of the min-sum algorithm when applied to the corresponding factor graph models of packing and covering LPs. We compare the solutions computed by the min-sum algorithm for packing and covering problems to the optimal solutions of the corresponding linear programming (LP) relaxations. In particular, we prove that if the LP has an optimal fractional solution, then for each fractional component, the min-sum algorithm either computes multiple solutions or the solution oscillates below and above the fraction. This implies that the min-sum algorithm computes the optimal integral solution only if the LP has a unique optimal solution that is integral. The converse is not true in general. For a special case of packing and covering problems, we prove that if the LP has a unique optimal solution that is integral and on the boundary of the box constraints, then the min-sum algorithm computes the optimal solution in pseudo-polynomial time. Our results unify and extend recent results for the maximum weight matching problem by [Sanghavi et al.,'2011] and [Bayati et al., 2011] and for the maximum weight independent set problem [Sanghavi et al.'2009].

Information Theory,Discrete Mathematics,Data Structures and Algorithms
Revisiting Degeneracy, Strict Feasibility, Stability, in Linear Programming

Jiyoung Im,Henry Wolkowicz

DOI: https://doi.org/10.48550/arXiv.2203.02795

2023-01-09

Abstract:Currently, the simplex method and the interior point method are indisputably the most popular algorithms for solving linear programs, LPs. Unlike general conic programs, LPs with a finite optimal value do not require strict feasibility in order to establish strong duality. Hence strict feasibility is seldom a concern, even though strict feasibility is equivalent to stability and a compact dual optimal set. This lack of concern is also true for other types of degeneracy of basic feasible solutions in LP. In this paper we discuss that the specific degeneracy that arises from lack of strict feasibility necessarily causes difficulties in both simplex and interior point methods. In particular, we show that the lack of strict feasibility implies that every basic feasible solution, BFS, is degenerate; thus conversely, the existence of a nondegenerate BFS implies that strict feasibility (regularity) holds. We prove the results using facial reduction and simple linear algebra. In particular, the facially reduced system reveals the implicit non-surjectivity of the linear map of the equality constraint system. As a consequence, we emphasize that facial reduction involves two steps where, the first guarantees strict feasibility, and the second recovers full row rank of the constraint matrix. This illustrates the implicit singularity of problems where strict feasibility fails, and also helps in obtaining new efficient techniques for preproccessing. We include an efficient preprocessing method that can be performed as an extension of phase-I of the two-phase simplex method. We show that this can be used to avoid the loss of precision for many well known problem sets in the literature, e.g., the NETLIB problem set.

Optimization and Control
Approximating rectangles by juntas and weakly-exponential lower bounds for LP relaxations of CSPs

Pravesh K. Kothari,Raghu Meka,Prasad Raghavendra

DOI: https://doi.org/10.1145/3055399.3055438

2017-06-19

Abstract:We show that for constraint satisfaction problems (CSPs), sub-exponential size linear programming relaxations are as powerful as nΩ(1)-rounds of the Sherali-Adams linear programming hierarchy. As a corollary, we obtain sub-exponential size lower bounds for linear programming relaxations that beat random guessing for many CSPs such as MAX-CUT and MAX-3SAT. This is a nearly-exponential improvement over previous results; previously, the best known lower bounds were quasi-polynomial in n (Chan, Lee, Raghavendra, Steurer 2013). Our bounds are obtained by exploiting and extending the recent progress in communication complexity for "lifting" query lower bounds to communication problems. The main ingredient in our results is a new structural result on "high-entropy rectangles" that may of independent interest in communication complexity.
Revisiting Tardos's Framework for Linear Programming: Faster Exact Solutions using Approximate Solvers

Daniel Dadush,Bento Natura,László A. Végh

DOI: https://doi.org/10.48550/arXiv.2009.04942

2020-09-10

Abstract:In breakthrough work, Tardos (Oper. Res. '86) gave a proximity based framework for solving linear programming (LP) in time depending only on the constraint matrix in the bit complexity model. In Tardos's framework, one reduces solving the LP $\min \langle c,{x}\rangle$, $Ax=b$, $x \geq 0$, $A \in \mathbb{Z}^{m \times n}$, to solving $O(nm)$ LPs in $A$ having small integer coefficient objectives and right-hand sides using any exact LP algorithm. This gives rise to an LP algorithm in time poly$(n,m\log\Delta_A)$, where $\Delta_A$ is the largest subdeterminant of $A$. A significant extension to the real model of computation was given by Vavasis and Ye (Math. Prog. '96), giving a specialized interior point method that runs in time poly$(n,m,\log\bar\chi_A)$, depending on Stewart's $\bar{\chi}_A$, a well-studied condition number. In this work, we extend Tardos's original framework to obtain such a running time dependence. In particular, we replace the exact LP solves with approximate ones, enabling us to directly leverage the tremendous recent algorithmic progress for approximate linear programming. More precisely, we show that the fundamental "accuracy" needed to exactly solve any LP in $A$ is inverse polynomial in $n$ and $\log\bar{\chi}_A$. Plugging in the recent algorithm of van den Brand (SODA '20), our method computes an optimal primal and dual solution using ${O}(m n^{\omega+1} \log (n)\log(\bar{\chi}_A+n))$ arithmetic operations, outperforming the specialized interior point method of Vavasis and Ye and its recent improvement by Dadush et al (STOC '20). At a technical level, our framework combines together approximate LP solutions to compute exact ones, making use of constructive proximity theorems -- which bound the distance between solutions of "nearby" LPs -- to keep the required accuracy low.

Optimization and Control,Data Structures and Algorithms
Graph Limits and Spectral Extremal Problems for Graphs

Lele Liu

DOI: https://doi.org/10.1137/22m1508807

IF: 0.868

2024-02-02

SIAM Journal on Discrete Mathematics

Abstract:SIAM Journal on Discrete Mathematics, Volume 38, Issue 1, Page 590-608, March 2024. We prove two conjectures in spectral extremal graph theory involving the linear combinations of graph eigenvalues. Let [math] be the largest eigenvalue of the adjacency matrix of a graph [math] and [math] be the complement of [math]. A nice conjecture states that the graph on [math] vertices maximizing [math] is the join of a clique and an independent set with [math] and [math] (also [math] and [math] if [math]) vertices, respectively. We resolve this conjecture for sufficiently large [math] using analytic methods. Our second result concerns the [math]-spread of a graph [math], which is defined as the difference between the largest eigenvalue and least eigenvalue of the signless Laplacian of [math]. It was conjectured by Cvetković, Rowlinson, and Simić [Publ. Inst. Math., 81 (2007), pp. 11–27] that the unique [math]-vertex connected graph of maximum [math]-spread is the graph formed by adding a pendant edge to [math]. We confirm this conjecture for sufficiently large [math].

mathematics, applied
Solving linear programs with complementarity constraints using branch-and-cut

Bin Yu,John E. Mitchell,Jong-Shi Pang

DOI: https://doi.org/10.1007/s12532-018-0149-2

2018-09-27

Mathematical Programming Computation

Abstract:<h3 class="a-plus-plus">Abstract</h3> <p class="a-plus-plus">A linear program with linear complementarity constraints (LPCC) requires the minimization of a linear objective over a set of linear constraints together with additional linear complementarity constraints. This class has emerged as a modeling paradigm for a broad collection of problems, including bilevel programs, Stackelberg games, inverse quadratic programs, and problems involving equilibrium constraints. The presence of the complementarity constraints results in a nonconvex optimization problem. We develop a branch-and-cut algorithm to find a global optimum for this class of optimization problems, where we branch directly on complementarities. We develop branching rules and feasibility recovery procedures and demonstrate their computational effectiveness in a comparison with CPLEX. The implementation builds on CPLEX through the use of callback routines. The computational results show that our approach is a strong alternative to constructing an integer programming formulation using big-<em class="a-plus-plus">M</em> terms to represent bounds for variables, with testing conducted on general LPCCs as well as on instances generated from bilevel programs with convex quadratic lower level problems.</p>
Linear Programming Heuristics for the Graph Isomorphism Problem

Reza Takapoui,Stephen Boyd

DOI: https://doi.org/10.48550/arXiv.1611.00711

2016-11-02

Optimization and Control

Abstract:An isomorphism between two graphs is a bijection between their vertices that preserves the edges. We consider the problem of determining whether two finite undirected weighted graphs are isomorphic, and finding an isomorphism relating them if the answer is positive. In this paper we introduce effective probabilistic linear programming (LP) heuristics to solve the graph isomorphism problem. We motivate our heuristics by showing guarantees under some conditions, and present numerical experiments that show effectiveness of these heuristics in the general case.
Linear Extension of the Erdos-Heilbronn Conjecture

Zhi‐Wei Sun,Liang Zhao

DOI: https://doi.org/10.48550/arxiv.0810.0467

2008-01-01

Abstract:The famous Erdos-Heilbronn conjecture plays an important role in the development of additive combinatorics. In 2007 Z. W. Sun made the following further conjecture (which is the linear extension of the Erdos-Heilbronn conjecture): For any finite subset A of a field F and nonzero elements $a_1,...,a_n$ of F, the set {a_1x_1+...+a_nx_n: x_1,....,x_n are distinct elements of A} has cardinality at least min{p(F)-delta, n(|A|-n)+1}, where the additive order p(F) of the multiplicative identity of F is different from n+1, and delta=0,1 takes the value 1 if and only if n=2 and $a_1+a_2=0$. In this paper we prove this conjecture of Sun when $p(F)\geq n(3n-5)/2$. We also obtain a sharp lower bound for the cardinality of the restricted sumset {x_1+...+x_n: x_1\in A_1,...,x_n\in A_n, and P(x_1,...,x_n)\not=0}, where $A_1,...,A_n$ are finite subsets of a field F and $P(x_1,...,x_n)$ is a general polynomial over F.
On the Circuit Diameter Conjecture for Counterexamples to the Hirsch Conjecture

Alexander E. Black,Steffen Borgwardt,Matthias Brugger

2024-04-09

Abstract:Circuit diameters of polyhedra are a fundamental tool for studying the complexity of circuit augmentation schemes for linear programming and for finding lower bounds on combinatorial diameters. The main open problem in this area is the circuit diameter conjecture, the analogue of the Hirsch conjecture in the circuit setting. A natural question is whether the well-known counterexamples to the Hirsch conjecture carry over. Previously, Stephen and Yusun showed that the Klee-Walkup counterexample to the unbounded Hirsch conjecture does not transfer to the circuit setting. Our main contribution is to show that the original counterexamples for the other variants, for bounded polytopes and using monotone walks, also do not transfer. Our results rely on new observations on structural properties of these counterexamples. To resolve the bounded case, we exploit the geometry of certain $2$-faces of the polytopes underlying all known bounded Hirsch counterexamples in Santos' work. For Todd's monotone Hirsch counterexample, we provide two alternative approaches. The first one uses sign-compatible circuit walks, and the second one uses the observation that Todd's polytope is anti-blocking. Along the way, we enumerate all linear programs over the polytope and find four new orientations that contradict the monotone Hirsch conjecture, while the remaining $7107$ satisfy the bound.

Combinatorics,Optimization and Control
A Stronger LP Bound for Formula Size Lower Bounds via Clique Constraints

Kenya Ueno

DOI: https://doi.org/10.48550/arXiv.0902.2146

2009-02-13

Abstract:We introduce a new technique proving formula size lower bounds based on the linear programming bound originally introduced by Karchmer, Kushilevitz and Nisan [11] and the theory of stable set polytope. We apply it to majority functions and prove their formula size lower bounds improved from the classical result of Khrapchenko [13]. Moreover, we introduce a notion of unbalanced recursive ternary majority functions motivated by a decomposition theory of monotone self-dual functions and give integrally matching upper and lower bounds of their formula size. We also show monotone formula size lower bounds of balanced recursive ternary majority functions improved from the quantum adversary bound of Laplante, Lee and Szegedy [15].

Computational Complexity

Refuting conjectures in extremal combinatorics via linear programming

On modeling hard combinatorial optimization problems as linear programs: Refutations of the "unconditional impossibility" claims

Linear Extension of the Erdős-Heilbronn Conjecture.

A Combinatorial Certifying Algorithm for Linear Programming Problems with Gainfree Leontief Substitution Systems

Some Zero-One Linear Programming Reformulations for the Maximum Clique Problem

Sparsified Linear Programming for Zero-Sum Equilibrium Finding

Three conjectures in extremal spectral graph theory

Zero-Sum Games and Linear Programming Duality

On the Power of Symmetric Linear Programs

0-1 Integer Linear Programming with a Linear Number of Constraints

Conic programming to understand sums of squares of eigenvalues of graphs

Analysis of the Min-Sum Algorithm for Packing and Covering Problems via Linear Programming

Revisiting Degeneracy, Strict Feasibility, Stability, in Linear Programming

Approximating rectangles by juntas and weakly-exponential lower bounds for LP relaxations of CSPs

Revisiting Tardos's Framework for Linear Programming: Faster Exact Solutions using Approximate Solvers

Graph Limits and Spectral Extremal Problems for Graphs

Solving linear programs with complementarity constraints using branch-and-cut

Linear Programming Heuristics for the Graph Isomorphism Problem

Linear Extension of the Erdos-Heilbronn Conjecture

On the Circuit Diameter Conjecture for Counterexamples to the Hirsch Conjecture

A Stronger LP Bound for Formula Size Lower Bounds via Clique Constraints