Abstract
We show the existence of a net near the sphere, such that the values of any matrix on the sphere and on the net are compared via a regularized Hilbert-Schmidt norm, which we introduce. This allows to construct an efficient net which controls the length of Ax for any random matrix A with independent columns (no other assumptions are required).
As a consequence we show that the smallest singular value σn (A) of an N × n random matrix A with i.i.d. mean zero, variance one entries enjoys the following small ball estimate, for any ϵ > 0
The proof of this result requires working with matrices whose rows are not independent, and, therefore, the fact that the theorem about discretization works for matrices with dependent rows, is crucial.
Furthermore, in the case of the square n×n matrix A with independent entries having concentration function separated from 1, i.i.d. rows, and such that \(\mathbb{E}\left\| A \right\|_{HS}^2 \le c{n^2}\), one has
for any ϵ > 0. In addition, for \(\epsilon > {c \over {\sqrt n }}\) the assumption of i.i.d. rows is not required. Our estimates generalize the previous results of Rudelson and Vershynin [29], [30], which required the sub-gaussian mean zero variance one assumptions, as well as the work of Rebrova and Tikhomirov [25], where mean zero variance 1 and i.i.d. entries were required.
Similar content being viewed by others
References
N. Alon and B. Klartag, Optimal compression of approximate inner products and dimension reduction, in 58th Annual IEEE Symposium on Foundations of Computer Science—FOCS 2017, IEEE Computer Society, Los Alamitos, CA, 2017, pp. 639–650.
Z. D. Bai and Y. Q. Yin, Necessary and sufficient conditions for almost sure convergence of the largest eigenvalue of a Wigner matrix, Ann. Probab. 16 (1988), 1729–1741.
Z. D. Bai and Y. Q. Yin, Limit of the smallest eigenvalue of a large-dimensional sample covariance matrix, Ann. Probab. 21 (1993), 1275–1294.
J. Beck, Irregularities of distribution, I, Acta Math. 159 (1987), 1–49.
J. Bourgain, V. H. Vu and P. M. Wood, On the singularity probability of discrete random matrices, J. Funct. Anal. 258 (2010), 559–603.
N. Cook, Lower bounds for the smallest singular value of structured random matrices, Ann. Probab. 46 (2018), 3442–3500.
A. Edelman, Eigenvalues and condition numbers of random matrices, SIAM J. Matrix Anal. Appl. 9 (1988), 543–560.
O. N. Feldheim and S. Sodin, A universality result for the smallest eigenvalues of certain sample covariance matrices, Geom. Funct. Anal. 20 (2010), 88–123.
W. Feller, A limit theorem for random variables with infinite moments, Amer. J. Math. 68 (1946), 257–262.
Y. Gordon, Some inequalities for Gaussian processes and applications, Israel J. Math. 50 (1985), 265–289.
O. Guedon, A. Litvak and K. Tatarko, Random polytopes obtained by matrices with heavy tailed entries, Commun. Contemp. Math. 22 (2020), Article no. 1950027.
R. Van Handel, R. Latala and P. Youssef, The dimension-free structure of nonhomogeneous random matrices, Invent. Math. 214 (2018), 1031–1080.
J. Kahn, J. Komlos and E. Szemeredi, On the probability that a random ±1 matrix is singular, J. Amer. Math. Soc. 8 (1995), 223–240.
R. Kannan and S. Vempala, Sampling lattice points, in Proceedings of the 29th ACM Symposium on the Theory of Computing (STOC’ 97), El Paso, ACM, NY, 1999, pp. 696–700.
B. Klartag and G. V. Livshyts, The lower bound for Koldobsky’s slicing inequality via random rounding, in Geometric Aspects of Functional Analysis, Springer, Cham, 2020, pp. 43–63.
V. Koltchinskii and S. Mendelson, Bounding the smallest singular value of a random matrix without concentration, Int. Math. Res. Not. IMRN 2015 (2015), 12991–13008.
R. Latala, Some estimates of norms of random matrices, Proc. Amer. Math. Soc. 133 (2005), 1273–1282.
J. E. Littlewood and A. C. Offord, On the number of real roots of a random algebraic equation. III, Rec. Math. [Mat. Sbornik] N. S. 12 (1943), 277–286.
A. Litvak and S. Spektor, Quantitative version of a Silverstein’s result, in Geometric Aspects of Functional Analysis, Springer, Cham, 2014, pp. 335–340.
G. Lugosi and S. Mendelson, Risk minimization by median-of-means tournaments, J. Eur. Math. Soc. (JEMS) 22 (2020), 925–965.
A. Litvak, A. Pajor, M. Rudelson and N. Tomczak-Jaegermann, Smallest singular value of random matrices and geometry of random polytopes, J. Reine Angew. Math. 589 (2005), 1–19.
A. Lytova and K. Tikhomirov, On delocalization of eigenvectors of random non-Hermitian matrices, Probab. Theory Related Fields 177 (2020), 465–524.
S. Mendelson and G. Paouris, On the singular values of random matrices, J. Eur. Math. Soc. (JEMS) 16 (2014), 823–834.
J. von Neumann and H. H. Goldstine, Numerical inverting of matrices of high order, Bull. Amer. Math. Soc. 53 (1947), 1021–1099.
E. Rebrova and K. Tikhomirov, Coverings of random ellipsoids, and invertibility of matrices with i.i.d. heavy-tailed entries, Israel J. of Math. 227 (2018), 507–544.
B. A. Rogozin, An estimate for the maximum of the convolution of bounded densities, Teor. Veroyatnost. i Primenen. 32 (1987), 53–61; English translation: Theory Probab. Appl. 32 (1987), 48–56.
M. Rudelson, Invertibility of random matrices: norm of the inverse, Ann. of Math. (2) 168 (2008), 575–600.
M. Rudelson, Delocalization of eigenvectors of random matrices, in Random Matrices, American Mathematical Society, Providence, RI, 2019, pp. 303–340.
M. Rudelson and R. Vershynin, The Littlewood-Offord problem and invertibility of random matrices, Adv. Math. 218 (2008), 600–633.
M. Rudelson and R. Vershynin, Smallest singular value of a random rectangular matrix, Comm. Pure Appl. Math. 62 (2009), 1707–1739.
M. Rudelson and R. Vershynin, Non-asymptotic theory of random matrices: extreme singular values, in Proceedings of the International Congress of Mathematicians. Vol. III, Hindustan Book Agency, New Delhi, 2010, pp. 1576–1602.
M. Rudelson and R. Vershynin, Small ball probabilities for linear images of high dimensional distributions, Int. Math. Res. Not. IMRN 19 (2015), 9594–9617.
M. Rudelson and R. Vershynin, Delocalization of eigenvectors of random matrices with independent entries, Duke Math. J. 164 (2015), 2507–2538.
J. Silverstein, On the weak limit of the largest eigenvalue of a large dimensional sample covariance matrix, J. Multivariate Anal. 30 (1989), 307–311.
A. Srinivasan, Approximation algorithms via randomized rounding: a survey, in Lectures on Approximation and Randomized Algorithms, Polish Scientific Publishers PWN, Warsaw, 1999, pp. 9–71.
S. Szarek, Condition numbers of random matrices, J. Complexity 7 (1991), 131–149.
T. Tao and V. Vu, On random ±1 matrices: Singularity and Determinant, Random Structures Algorithms 28 (2006), 1–23.
T. Tao and V. Vu, On the singularity probability of random Bernoulli matrices, J. Amer. Math. Soc. 20 (2007), 603–628.
T. Tao and V. Vu, Inverse Littlewood-Offord theorems and the condition number of random discrete matrices, Ann. of Math. (2) 169 (2009), 595–632.
T. Tao and V. Vu, Random matrices: the distribution of the smallest singular values, Geom. Funct. Anal. 20 (2010), 260–297.
K. Tatarko, An upper bound on the smallest singular value of a square random matrix, J. Complexity 48 (2018), 119–128.
K. Tikhomirov, The limit of the smallest singular value of random matrices with i.i.d. entries, Adv. Math. 284 (2015), 1–20.
K. Tikhomirov, The smallest singular value of random rectangular matrices with no moment assumptions on entries, Israel J. Math. 212 (2016), 289–314.
K. Tikhomirov, Invertibility via distance for non-centered random matrices with continuous distributions, Random Structures Algorithms 57 (2020), 526–562.
K. Tikhomirov, Singularity of random Bernoulli matrices, Ann. of Math. (2) 191 (2020), 593–634.
R. Vershynin, Spectral norm of products of random and deterministic matrices, Probab. Theory Related Fields 150 (2011), 471–509.
R. Vershynin, High-Dimensional Probability, Cambridge University Press, Cambridge, 2018.
Acknowledgements
First of all, the author thanks Bo’az Klartag for his support, encouragement, enduring patience and inspiration during the time this paper was written; this project sprouted as a follow-up to our joint work [15], however Bo’az decided not to join it at this point. In addition, the author would like to thank Dr. Klartag for mentoring her during the Fall 2017 MSRI program, which has led to a phase transition in her knowledge, understanding and proficiency in the subject.
The author is very grateful to Mark Rudelson for encouragement and helpful discussions, which have led to a significant improvement of her understanding of the field in general. The author is grateful to Konstantin Tikhomirov for helpful discussions, and in particular for relating to her what types of behaviors of the smallest singular values are expected by the experts. The author is grateful to Roman Vershynin for bringing to her attention the question about matrices whose entries have different moments.
The author also thanks to the anonymous referee for many valuable comments which helped to significantly improve the presentation and motivated further exploration.
The author is supported in part by the NSF CAREERDMS-1753260. The work was partially supported by the National Science Foundation under Grant No. DMS-1440140 while the author was in residence at the Mathematical Sciences Research Institute in Berkeley, California, during the Fall 2017 semester.
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Livshyts, G.V. The smallest singular value of heavy-tailed not necessarily i.i.d. random matrices via random rounding. JAMA 145, 257–306 (2021). https://doi.org/10.1007/s11854-021-0183-2
Received:
Revised:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11854-021-0183-2