Abstract
In this article, we disprove a conjecture of Goemans and Linial; namely, that every negative type metric embeds into ℓ1 with constant distortion. We show that for an arbitrarily small constant δ > 0, for all large enough n, there is an n-point negative type metric which requires distortion at least (log log n)1/6-δ to embed into ℓ1. Surprisingly, our construction is inspired by the Unique Games Conjecture (UGC), establishing a previously unsuspected connection between probabilistically checkable proof systems (PCPs) and the theory of metric embeddings. We first prove that the UGC implies a super-constant hardness result for the (nonuniform) SPARSESTCUT problem. Though this hardness result relies on the UGC, we demonstrate, nevertheless, that the corresponding PCP reduction can be used to construct an “integrality gap instance” for SPARSESTCUT. Towards this, we first construct an integrality gap instance for a natural SDP relaxation of UNIQUEGAMES. Then we “simulate” the PCP reduction and “translate” the integrality gap instance of UNIQUEGAMES to an integrality gap instance of SPARSESTCUT. This enables us to prove a (log log n)1/6-δ integrality gap for SPARSESTCUT, which is known to be equivalent to the metric embedding lower bound.
- Amit Agarwal, Moses Charikar, Konstantin Makarychev, and Yury Makarychev. 2005. O(√ log n) approximation algorithms for min-UnCut, min-2CNF deletion, and directed cut problems. In Proceedings of the ACM Symposium on the Theory of Computing. 573--581. Google ScholarDigital Library
- Sanjeev Arora, Boaz Barak, and David Steurer. 2010. Subexponential algorithms for unique games and related problems. In Proceedings of FOCS. 563--572. Google ScholarDigital Library
- Sanjeev Arora, Eli Berger, Elad Hazan, Guy Kindler, and Muli Safra. 2005. On non-approximability for quadratic programs. In Proceedings of FOCS. 206--215. Google ScholarDigital Library
- Sanjeev Arora, Subhash Khot, Alexandra Kolla, David Steurer, Madhur Tulsiani, and Nisheeth K. Vishnoi. 2008. Unique games on expanding constraint graphs are easy. In Proceedings of STOC. 21--28. Google ScholarDigital Library
- Sanjeev Arora, James R. Lee, and Assaf Naor. 2007. Fréchet embeddings of negative type metrics. Disc. Computat. Geom. 38, 4, 726--739. Google ScholarDigital Library
- Sanjeev Arora, Satish Rao, and Umesh V. Vazirani. 2009. Expander flows, geometric embeddings and graph partitioning. J. ACM 56, 2. Google ScholarDigital Library
- Yonatan Aumann and Yuval Rabani. 1998. An O(log k) approximate min-cut max-flow theorem and approximation algorithm. SIAM J. Comput. 27, 1, 291--301. Google ScholarDigital Library
- Tim Austin, Assaf Naor, and Alain Valette. 2010. The Euclidean distortion of the lamplighter group. Disc. Computat. Geom. 44, 1, 55--74.Google ScholarDigital Library
- Boaz Barak, Parikshit Gopalan, Johan Håstad, Raghu Meka, Prasad Raghavendra, and David Steurer. 2012. Making the long code shorter. In Proceedings of FOCS. 370--379. Google ScholarDigital Library
- Mihir Bellare, Oded Goldreich, and Madhu Sudan. 1998. Free bits, PCPs, and non-approximability -- towards tight results. In SIAM J. Comput. 422--431. Google ScholarDigital Library
- Jean Bourgain. 1985. On Lipschitz embeddings of finite metric spaces in Hilbert space. Israel J. Math. 52, 46--52.Google ScholarCross Ref
- Jean Bourgain. 2002. On the distribution of the Fourier spectrum of Boolean functions. Isr. J. Math. 131, 269--276.Google ScholarCross Ref
- Shuchi Chawla, Anupam Gupta, and Harald Räcke. 2008. Embeddings of negative-type metrics and an improved approximation to generalized sparsest cut. ACM Trans. Algor. 4, 2. Google ScholarDigital Library
- Shuchi Chawla, Robert Krauthgamer, Ravi Kumar, Yuval Rabani, and D. Sivakumar. 2006. On the hardness of approximating multicut and sparsest-cut. Computat. Complex. 15, 2, 94--114. Google ScholarDigital Library
- Jeff Cheeger and Bruce Kleiner. 2010. Differentiating maps into L1, and the geometry of BV functions. Ann. Math. 171, 1347--1385.Google ScholarCross Ref
- Jeff Cheeger, Bruce Kleiner, and Assaf Naor. 2009. A (log n)Ω(1) integrality gap for the sparsest cut SDP. In Proceedings of FOCS. 555--564. Google ScholarDigital Library
- Nikhil R. Devanur, Subhash Khot, Rishi Saket, and Nisheeth K. Vishnoi. 2006. Integrality gaps for sparsest cut and minimum linear arrangement problems. In Proceedings of STOC. 537--546. Google ScholarDigital Library
- M. Deza and Monique Laurent. 1997. Geometry of Cuts and Metrics. Springer-Verlag, New York. Google ScholarDigital Library
- Peter Enflo. 1969. On the non-existence of uniform homeomorphism between Lp spaces. Arkiv. Mat. 8, 103--105.Google ScholarCross Ref
- Uriel Feige, MohammadTaghi Hajiaghayi, and James R. Lee. 2008. Improved approximation algorithms for minimum weight vertex separators. SIAM J. Comput. 38, 2, 629--657. Google ScholarDigital Library
- Uriel Feige and László Lovász. 2002. Two-prover one-round proof systems, their power and their problems. In Proceedings of the ACM Symposium on the Theory of Computing. 733--744. Google ScholarDigital Library
- Uriel Feige and Gideon Schechtman. 2002. On the optimality of the random hyperplane rounding technique for MAX CUT. Random Struct. Algor. 20, 3, 403--440. Google ScholarDigital Library
- Michel X. Goemans. 1997. Semidefinite programming in combinatorial optimization. Math. Program. 79, 143--161. Google ScholarDigital Library
- Michel X. Goemans and David P. Williamson. 1995. Improved approximation algorithms for maximum cut and satisfiability problems using semidefinite programming. J. ACM 42, 6, 1115--1145. Google ScholarDigital Library
- Jeff Kahn, Gil Kalai, and Nathan Linial. 1988. The influence of variables on Boolean functions (extended abstract). In Proceedings of FOCS. 68--80. Google ScholarDigital Library
- Daniel M. Kane and Raghu Meka. 2013. A PRG for Lipschitz functions of polynomials with applications to sparsest cut. In Proceedings of the ACM Symposium on the Theory of Computing. Google ScholarDigital Library
- George Karakostas. 2009. A better approximation ratio for the vertex cover problem. ACM Trans. Algor. 5, 4. Google ScholarDigital Library
- Subhash Khot. 2002. On the power of unique 2-prover 1-round games. In Proceedings of the ACM Symposium on the Theory of Computing. 767--775. Google ScholarDigital Library
- Subhash Khot and Assaf Naor. 2006. Nonembeddability theorems via Fourier analysis. Math. Ann. 334, 4, 821--852. DOI:http://dx.doi.org/10.1007/s00208-005-0745-0Google ScholarCross Ref
- Subhash Khot and Ryan O'Donnell. 2009. SDP gaps and UGC-hardness for Max-Cut-Gain. Theory Comput. 5, 1, 83--117.Google ScholarCross Ref
- Subhash Khot and Rishi Saket. 2009. SDP integrality gaps with local ℓ1-embeddability. In Proceedings of FOCS. 565--574. Google ScholarDigital Library
- Subhash Khot and Nisheeth K. Vishnoi. 2005. The unique games conjecture, integrality gap for cut problems and embeddability of negative type metrics into ℓ1. In Proceedings of FOCS. 53--62. Google ScholarDigital Library
- Robert Krauthgamer and Yuval Rabani. 2009. Improved lower bounds for embeddings into L1. SIAM J. Comput. 38, 6, 2487--2498. Google ScholarDigital Library
- James R. Lee and Assaf Naor. 2006. Lp metrics on the Heisenberg group and the Goemans-Linial conjecture. In Proceedings of FOCS. 99--108. Google ScholarDigital Library
- Frank Thomson Leighton and Satish Rao. 1999. Multicommodity max-flow min-cut theorems and their use in designing approximation algorithms. J. ACM 46, 6, 787--832. Google ScholarDigital Library
- Nathan Linial. 2002. Finite metric spaces: Combinatorics, geometry and algorithms. In Proceedings of the International Congress of Mathematicians. 573--586.Google ScholarDigital Library
- Nathan Linial, Eran London, and Yuri Rabinovich. 1995. The geometry of graphs and some of its algorithmic applications. Combinatorica 15, 2, 215--245.Google ScholarCross Ref
- Jiri Matousek. 2002. Lectures on Discrete Geometry. Graduate Texts in Mathematics, vol. 212. Springer-Verlag. Google ScholarDigital Library
- Assaf Naor, Yuval Rabani, and Alistair Sinclair. 2005. Quasisymmetric embeddings, the observable diameter, and expansion properties of graphs. J. Funct. Anal. 227, 2, 273--303.Google ScholarCross Ref
- R. O'Donnell. 2004. Computational aspects of noise sensitivity. Ph.D. Dissertation. MIT. Google ScholarDigital Library
- Prasad Raghavendra. 2008. Optimal algorithms and inapproximability results for every CSP? In Proceedings of STOC. 245--254. Google ScholarDigital Library
- Prasad Raghavendra and David Steurer. 2009. Integrality gaps for strong SDP relaxations of UNIQUE GAMES. In Proceedings of FOCS. 575--585. Google ScholarDigital Library
- Prasad Raghavendra and David Steurer. 2010. Graph expansion and the unique games conjecture. In Proceedings of STOC. 755--764. Google ScholarDigital Library
- Gideon Schechtman. 2003. Handbook of the Geometry of Banach Spaces. Vol. 2, North Holland, Chapter Concentration results and applications.Google Scholar
- Nisheeth K. Vishnoi. 2013. Lx = b. Foundations and Trends® in Theoretical Computer Science. Vol. 8, 1--2, 1--141. http://dx.doi.org/10.1561/0400000054.Google Scholar
Index Terms
- The Unique Games Conjecture, Integrality Gap for Cut Problems and Embeddability of Negative-Type Metrics into ℓ1
Recommendations
Graph expansion and the unique games conjecture
STOC '10: Proceedings of the forty-second ACM symposium on Theory of computingThe edge expansion of a subset of vertices S ⊆ V in a graph G measures the fraction of edges that leave S. In a d-regular graph, the edge expansion/conductance Φ(S) of a subset S ⊆ V is defined as Φ(S) = (|E(S, V\S)|)/(d|S|). Approximating the ...
Integrality Gaps for Strong SDP Relaxations of UNIQUE GAMES
FOCS '09: Proceedings of the 2009 50th Annual IEEE Symposium on Foundations of Computer ScienceWith the work of Khot and Vishnoi (FOCS 2005) as a starting point, we obtain integrality gaps for certain strong SDP relaxations of unique games. Specifically, we exhibit a gap instance for the basic semidefinite program strengthened by all valid linear ...
The Unique Games Conjecture, Integrality Gap for Cut Problems and Embeddability of Negative Type Metrics into \ell 1
FOCS '05: Proceedings of the 46th Annual IEEE Symposium on Foundations of Computer ScienceIn this paper we disprove the following conjecture due to Goemans [16] and Linial [24] (also see [5, 26]): "Every negative type metric embeds into \ell 1 with constant distortion." We show that for every \delta} \ge 0, and for large enough n, there is ...
Comments