Abstract
In this article, we investigate high-dimensional band sample covariance matrices under the regime that the sample size n, the dimension p, and the bandwidth d tend simultaneously to infinity such that
It is shown that the empirical spectral distribution of those matrices converges weakly to a deterministic probability measure with probability 1. The limiting measure is characterized by its moments. Certain restricted compositions of natural numbers play a crucial role in the evaluation of the expected moments of the empirical spectral distribution.
Similar content being viewed by others
References
Abramson, M.: Restricted combinations and compositions. Fibonacci Quart. 14, 439–452 (1976)
Bai, Z., Silverstein, J.: Spectral Analysis of Large Dimensional Random Matrices. Springer, New York (2010)
Bai, Z.D., Yin, Y.Q.: Limit of the smallest eigenvalue of a large-dimensional sample covariance matrix. Ann. Prob. 21, 1275–1294 (1993)
Bai, Z.D., Zhang, L.X.: Semicircle law for Hadamard products. SIAM. J. Matrix Anal. Appl. 29, 473–495 (2007)
Benaych-Georges, F., Péché, S.: Largest eigenvalues and eigenvectors of band or sparse matrices. Electron. Commun. Probab. 19, 1–9 (2014a)
Benaych-Georges, F., Péché, S.: Localization and delocalization for heavy tailed band matrices. Ann. Inst. Henri Poincaré Probab. Stat. 50, 1385–1403 (2014b)
Bickel, P., Levina, E.: Covariance regularization by thresholding. Ann. Stat. 36, 2577–2604 (2008a)
Bickel, P., Levina, E.: Regularized estimation of large covariance matrices. Ann. Stat. 36, 199–227 (2008b)
Bogachev, L.V., Molchanov, S.A., Pastur, L.A.: On the density of states of random band matrices. Mat. Zametki. 50, 31–42 (1991)
Cai, T., Jiang, T.: Limiting laws of coherence of random matrices with applications to testing covariance structure and construction of compressed sensing matrices. Ann. Stat. 39, 1496–1525 (2011)
Cai, T., Zhou, H.: Optimal rates of convergence for sparse covariance matrix estimation. Ann. Stat. 40, 2389–2420 (2012)
El Karoui, N.: Operator norm consistent estimation of large-dimensional sparse covariance matrices. Ann. Stat. 36, 2717–2756 (2008)
Hoeffding, W.: Probability inequalities for sums of bounded random variables. J. Am. Stat. Assoc. 58, 13–30 (1963)
Kirsch, W., Khorunzhy, A.: Limit of infinite band width for product of two random matrices. Random Oper. Stoch. Eq. 5, 325–336 (1997)
Lam, C., Fan, J.: Sparsistency and rates of convergence in large covariance matrix estimation. Ann. Stat. 37, 4254–4278 (2009)
Levina, E., Vershynin, R.: Partial estimation of covariance matrices. Prob. Theory Relat. Fields 153, 405–419 (2012)
Li, C.K., Mathias, R.: The Lidskii–Mirsky–Wielandt theorem—additive and multiplicative versions. Numer. Math. 81, 377–413 (1999)
Qiu, Y., Chen, S.: Test for bandedness of high-dimensional covariance matrices and bandwidth estimation. Ann. Stat. 40, 1285–1314 (2012)
Sodin, S.: The spectral edge of some random band matrices. Ann. Math. 172, 2223–2251 (2010)
Author information
Authors and Affiliations
Corresponding author
Additional information
Supported by the DFG Research Unit 1735, RO 3766/3-1 and DE 502/26-2.
Rights and permissions
About this article
Cite this article
Jurczak, K. Weak Convergence of the Empirical Spectral Distribution of High-Dimensional Band Sample Covariance Matrices. J Theor Probab 31, 1273–1302 (2018). https://doi.org/10.1007/s10959-017-0751-7
Received:
Revised:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10959-017-0751-7
Keywords
- High-dimensional sample covariance matrices
- Empirical spectral distribution
- Strong convergence
- Weak convergence
- Method of moments
- Number of restricted compositions of a natural number