Noninteractive correlation distillation, inhomogeneous Markov chains, and the reverse BonamiBeckner inequality
 199 Downloads
 27 Citations
Abstract
In this paper we studynoninteractive correlation distillation (NICD), a generalization of noise sensitivity previously considered in [5, 31, 39]. We extend the model toNICD on trees. In this model there is a fixed undirected tree with players at some of the nodes. One node is given a uniformly random string and this string is distributed throughout the network, with the edges of the tree acting as independent binary symmetric channels. The goal of the players is to agree on a shared random bit without communicating.

• In the case of akleaf star graph (the model considered in [31]), we resolve the open question of whether the success probability must go to zero ask » ∞. We show that this is indeed the case and provide matching upper and lower bounds on the asymptotically optimal rate (a slowlydecaying polynomial).

• In the case of thekvertex path graph, we show that it is always optimal for all players to use the same 1bit function.

• In the general case we show that all players should use monotone functions. We also show, somewhat surprisingly, that for certain trees it is better if not all players use the same function.
Our techniques include the use of thereverse BonamiBeckner inequality. Although the usual BonamiBeckner has been frequently used before, its reverse counterpart seems not to be well known. To demonstrate its strength, we use it to prove a new isoperimetric inequality for the discrete cube and a new result on the mixing of short random walks on the cube. Another tool that we need is a tight bound on the probability that a Markov chain stays inside certain sets; we prove a new theorem generalizing and strengthening previous such bounds [2, 3, 6]. On the probabilistic side, we use the “reflection principle” and the FKG and related inequalities in order to study the problem on general trees.
Keywords
Markov Chain Boolean Function Isoperimetric Inequality Optimal Protocol Random StringPreview
Unable to display preview. Download preview PDF.
References
 [1]M. Abramowitz and I. Stegun,Handbook of Mathematical Functions, Dover, New York, 1972.zbMATHGoogle Scholar
 [2]M. Ajtai, J. Komolós and E. Szemerédi,Deterministic simulation in LOGSPACE, inProceedings of the 19th Annual ACM Symposium on Theory of Computing, ACM, New York, 1987, pp. 132–140.Google Scholar
 [3]N. Alon, U. Feige, A. Wigderson and D. Zuckerman,Derandomized graph products, Computational Complexity, Birkhäuser, Basel, 1995.Google Scholar
 [4]N. Alon, G. Kalai, M. Ricklin and L. Stockmeyer,Lower bounds on the competitive ratio for mobile user tracking and distributed job scheduling, Theoretical Computer Science130 (1994), 175–201.zbMATHCrossRefMathSciNetGoogle Scholar
 [5]N. Alon, U. Maurer and A. Wigderson, Unpublished results, 1991.Google Scholar
 [6]N. Alon and J. Spencer,The Probabilistic Method, 2nd edn., Wiley, New York, 2000.zbMATHGoogle Scholar
 [7]K. Amano and A. Maruoka,On learning monotone Boolean functions under the uniform distribution, Lecture Notes in Computer Science2533, Springer, New York, 2002, pp. 57–68.Google Scholar
 [8]W. Beckner,Ineqalities in Fourier analysis, Annals of Mathematics102 (1975), 159–182.CrossRefMathSciNetGoogle Scholar
 [9]M. BenOr and N. Linial,Collective coin flipping, inRandomness and Computation (S. Micali, ed.), Academic Press, New York, 1990.Google Scholar
 [10]I. Benjamini, G. Kalai and O Schramm,Noise sensitivity of boolean functions and applications to percolation, Publications Mathématiques de l'Institut des Hautes Études Scientifiques90 (1999), 5–43.zbMATHCrossRefMathSciNetGoogle Scholar
 [11]S. Bobkov and F. Götze,Discrete isoperimetric, and Poincarétype inequalities, Probability theory and Related Fields114 (1999), 245–277.zbMATHCrossRefMathSciNetGoogle Scholar
 [12]A. Bonami,Études des coefficients Fourier des fonctiones de L ^{p} (G), Annales de l'Institut Fourier20 (1970), 335–402.zbMATHMathSciNetGoogle Scholar
 [13]C. Borell,Positivity improving operators and hypercontractivity, Mathematische Zeitschrift180 (1982), 225–234.zbMATHCrossRefMathSciNetGoogle Scholar
 [14]J. Bourgain, An appendix toSharp thresholds of graph properties, and the ksat problem, by E. Friedgut, Journal of the American Mathematical Society12 (1999), 1017–1054.zbMATHCrossRefMathSciNetGoogle Scholar
 [15]J. Bourgain, J. Kahn, G. Kalai, Y. Katznelson and N. Linial,The influence of variables in product spaces. Israel Journal of Mathematics77 (1992), 55–64.zbMATHMathSciNetGoogle Scholar
 [16]J. Bourgain and G. Kalai,Influences of variables and threshold intervals under group symmetries, Geometric and Functional Analysis7 (1997), 438–461.zbMATHCrossRefMathSciNetGoogle Scholar
 [17]N. Bshouty, J. Jackson and C. Tamon,Uniformdistribution attribute noise learnability, inProceedings of the Eighth Annual Conference on Computational Learning Theory, (COLT 1995), Santa Cruz, California, USA, ACM, 1995.Google Scholar
 [18]I. Dinur, V. Guruswami and S. Khot,Vertex cover on kuniform hypergraphs is hard to approximate within factor (k3∈), ECCC Technical Report TRO2027, 2002.Google Scholar
 [19]I. Dinur and S. Safra,The importance of being biased, inProceedings of the 34th Annual ACM Symposium on the Theory of Computing, ACM, New York, 2002, pp. 33–42.Google Scholar
 [20]W. Feller,An Introduction to Probability Theory and its Applications, 3rd edn., Wiley, New York, 1968.zbMATHGoogle Scholar
 [21]C. Fortuin, P. Kasteleyn and J. Ginibre,Correlation inequalities on some partially ordered sets, Communications in Mathematical Physics22 (1971), 89–103.zbMATHCrossRefMathSciNetGoogle Scholar
 [22]E. Friedgut,Boolean functions with low average sensivity depend on few coordinates, Combinatorica18 (1998), 474–483.CrossRefMathSciNetGoogle Scholar
 [23]E. Friedgut and G. Kalai,Every monotone graph property has a sharp threshold, Proceedings of the American Mathematical Society124 (1996), 2993–3002.zbMATHCrossRefMathSciNetGoogle Scholar
 [24]H. O. Georgii,Gibbs Measures and Phase Transitions, Volume 9 of de Gruyter Studies in Mathematics, de Gruyter, Berlin, 1988.zbMATHGoogle Scholar
 [25]G. Hardy, J. Littlewood and G. Póyla,Inequalities, 2nd edn., Cambridge University Press, 1952.Google Scholar
 [26]J. H»stad,Some optimal inapproximability results, Journal of the ACM48 (2001), 798–869.CrossRefMathSciNetGoogle Scholar
 [27]J. Kahn, G. Kalai and N. LinialThe influence of variables on boolean functions, inProceedings of the 29th Annual IEEE Symposium on Foundations of Computer Science, IEEE Computer Society Press, Los Alamitos, CA, 1988, pp. 68–80.Google Scholar
 [28]S. Khot,On the power of unique 2prover 1round games, inProceedings of the 34th Annual ACM Symposium on the Theory of Computing, ACM, New York, 2002, pp. 767–775.Google Scholar
 [29]D. Kleitman,Families of nondisjoint subsets, Journal of Combinatorial Theory1 (1966), 153–155.zbMATHCrossRefMathSciNetGoogle Scholar
 [30]A. Klivans, R. O'Donnell and R. Servedio,Learning intersections and thresholds of halfspaces, inProceedings of the 43rd Annual IEEE Symposium on Foundations of Computer Science, IEEE Computer Society Press, Los Alamitos, CA, 2002, pp. 177–186.CrossRefGoogle Scholar
 [31]E. Mossel and R. O'Donnell,Coin flipping from a cosmic source: On error correction of truly random bits, Random Structures & Algorithms26 (2005), 418–436.zbMATHCrossRefMathSciNetGoogle Scholar
 [32]A. Naor, E. Friedgut and G. Kalai,Boolean functions whose Fourier transform is concentrated on the first two levels, Advances in Applied Mathematics29 (2002), 427–437.zbMATHCrossRefMathSciNetGoogle Scholar
 [33]R. O'Donnell,Hardness amplification within NP, inProceedings of the 34th Annual ACM Symposium on the Theory of Computing, ACM, New York, 2002, pp. 715–760.Google Scholar
 [34]R. O'Donnell,Computational applications of noise sensitivity, PhD thesis, Massachusetts Institute of Technology, 2003.Google Scholar
 [35]R. O'Donnell and R. Servedio,Learning monotone decision trees, Manuscript, 2004.Google Scholar
 [36]R. Raz,Fourier analysis for probabilistic communication complexity, Computational Complexity5 (1995), 205–221.zbMATHCrossRefMathSciNetGoogle Scholar
 [37]V. Sazonov,Normal Approximation— Some Recent Advances, SpringerVerlag, Berlin, 1981.zbMATHGoogle Scholar
 [38]M. Talagrand,On Russo's approximate 0–1 law, Annals of Probability22 (1994), 1476–1387.Google Scholar
 [39]K. Yang,On the (im)possibility of noninteractive correlation distillation, inLATIN 2004: Theoretical Informatics, 6th Latin American Symposium, Buenos Aires, Argentina, April 5–8, 2004, Proceedings (M. FarachColton, ed.), Lecture Notes in Computer Science 2976, Springer, Berlin, 2004.Google Scholar