Concentration and moment inequalities for polynomials of independent random variables

In this work we design a general method for proving moment inequalities for polynomials of independent random variables. Our method works for a wide range of random variables including Gaussian, Boolean, exponential, Poisson and many others. We apply our method to derive general concentration inequalities for polynomials of independent random variables. We show that our method implies concentration inequalities for some previously open problems, e.g. permanent of random symmetric matrices. We show that our concentration inequality is stronger than the well-known concentration inequality due to Kim and Vu [29]. The main advantage of our method in comparison with the existing ones is a wide range of random variables we can handle and bounds for previously intractable regimes of high degree polynomials and small expectations. On the negative side we show that even for boolean random variables each term in our concentration inequality is tight.

[1]  E. Giné,et al.  On decoupling, series expansions, and tail behavior of chaos processes , 1993 .

[2]  Shachar Lovett,et al.  Polynomial Threshold Functions: Structure, Approximation and Pseudorandomness , 2009, ArXiv.

[3]  M. Talagrand.  Concentration of measure and isoperimetric inequalities in product spaces , 1994, math/9406212.

[4]  Christos Faloutsos,et al.  DOULION: counting triangles in massive graphs with a coin , 2009, KDD.

[5]  V. Vu,et al.  On the permanent of random Bernoulli matrices , 2008 .

[6]  V. Vu.  On the Concentration of Multi-Variate Polynomials with Small Expectation , 2000 .

[7]  Prasad Raghavendra,et al.  Average Sensitivity and Noise Sensitivity of Polynomial Threshold Functions , 2009, SIAM J. Comput..

[8]  Stephen P. Boyd,et al.  Convex Optimization , 2004, Algorithms and Theory of Computation Handbook.

[9]  S. Janson.  Gaussian Hilbert Spaces , 1997 .

[10]  A. Carbery,et al.  Distributional and L-q norm inequalities for polynomials over convex bodies in R-n , 2001 .

[11]  Rados law,et al.  Tail and moment estimates for chaoses generated by symmetric random variables with logarithmically concave tails , 2011 .

[12]  Prahladh Harsha,et al.  Bounding the Sensitivity of Polynomial Threshold Functions , 2014, Theory Comput..

[13]  Claire Mathieu,et al.  Correlation clustering with noisy input , 2010, SODA '10.

[14]  Van H. Vu.  On the concentration of multivariate polynomials with small expectation , 1999, Random Struct. Algorithms.

[15]  Alexander Barvinok,et al.  Matrices with prescribed row and column sums , 2010, 1010.5706.

[16]  W. Feller.  Generalization of a probability limit theorem of Cramér , 1943 .

[17]  Prasad Raghavendra,et al.  Average Sensitivity and Noise Sensitivity of Polynomial Threshold Functions , 2014, SIAM J. Comput..

[18]  Rocco A. Servedio,et al.  A Regularity Lemma, and Low-Weight Approximators, for Low-Degree Polynomial Threshold Functions , 2009, 2010 IEEE 25th Annual Conference on Computational Complexity.

[19]  Rafal Latala,et al.  Tail and moment estimates for chaoses generated by symmetric random variables with logarithmically concave tails , 2010, 1007.1431.

[20]  M. Habib.  Probabilistic methods for algorithmic discrete mathematics , 1998 .

[21]  Noga Alon,et al.  The Probabilistic Method , 2015, Fundamentals of Ramsey Theory.

[22]  Gábor Lugosi,et al.  Concentration Inequalities , 2008, COLT.

[23]  Svante Janson,et al.  Poisson Approximation for Large Deviations , 1990, Random Struct. Algorithms.

[24]  Noga Alon,et al.  The Probabilistic Method, Third Edition , 2008, Wiley-Interscience series in discrete mathematics and optimization.

[25]  Rafa l Lata la,et al.  Moment and tail estimates for multidimensional chaoses generated by positive random variables with logarithmically concave tails , 2005 .

[26]  Fan Chung Graham,et al.  Concentration Inequalities and Martingale Inequalities: A Survey , 2006, Internet Math..

[27]  Mihail N. Kolountzakis,et al.  Approximate Triangle Counting , 2009, ArXiv.

[28]  M. Ledoux.  The concentration of measure phenomenon , 2001 .

[29]  Van H. Vu,et al.  Concentration of non‐Lipschitz functions and applications , 2002, Random Struct. Algorithms.

[30]  Van H. Vu,et al.  Concentration of Multivariate Polynomials and Its Applications , 2000, Comb..

[31]  Scott Aaronson,et al.  The computational complexity of linear optics , 2010, STOC '11.

[32]  Alessandro Panconesi,et al.  Concentration of Measure for the Analysis of Randomized Algorithms , 2009 .

[33]  Svante Janson,et al.  The infamous upper tail , 2002, Random Struct. Algorithms.

[34]  D. Nualart.  GAUSSIAN HILBERT SPACES (Cambridge Tracts in Mathematics 129) By SVANTE JANSON: 340 pp., £40.00, ISBN 0 521 56128 0 (Cambridge University Press, 1997) , 1998 .

[35]  Rados Law Adamczak,et al.  Logarithmic Sobolev inequalities and concentration of measure for convex functions and polynomial chaoses , 2005 .

[36]  M. Bagnoli,et al.  Log-concave probability and its applications , 2004 .

[37]  S. Boucheron,et al.  Moment inequalities for functions of independent random variables , 2005, math/0503651.

[38]  C. McDiarmid.  Concentration , 1862, The Dental register.

[39]  Rocco A. Servedio,et al.  Bounded Independence Fools Halfspaces , 2009, 2009 50th Annual IEEE Symposium on Foundations of Computer Science.

[40]  Prabhakar Raghavan,et al.  Randomized rounding: A technique for provably good algorithms and algorithmic proofs , 1985, Comb..

[41]  C. Donati-Martin,et al.  Moments of the Gaussian Chaos , 2012 .

[42]  Peter Major.  Estimates on the tail behavior of Gaussian polynomials. The discussion of a result of Latala , 2009 .

[43]  Rafa l Lata la.  Estimates of moments and tails of Gaussian chaoses , 2005 .

[44]  Michael Friendly,et al.  Discrete Distributions , 2005, Probability and Bayesian Modeling.

[45]  R. Latala.  Estimates of moments and tails of Gaussian chaoses , 2005, math/0505313.

[46]  M. An.  Log-Concave Probability Distributions: Theory and Statistical Testing , 1996 .

[47]  M. Talagrand.  A new look at independence , 1996 .

[48]  Rafał M. Łochowski.  Moment and tail estimates for multidimensional chaoses generated by symmetric random variables with logarithmically concave tails , 2006 .

[49]  Daniel M. Kane.  The Gaussian Surface Area and Noise Sensitivity of Degree-d Polynomial Threshold Functions , 2010, 2010 IEEE 25th Annual Conference on Computational Complexity.

[50]  F. T. Wright,et al.  A Bound on Tail Probabilities for Quadratic Forms in Independent Random Variables , 1971 .

[51]  David Zuckerman,et al.  Pseudorandom generators for polynomial threshold functions , 2009, STOC '10.