The Annals of Applied Probability

Local law and Tracy–Widom limit for sparse sample covariance matrices

Jong Yun Hwang, Ji Oon Lee, and Kevin Schnelli

Full-text: Access denied (no subscription detected)

We're sorry, but we are unable to provide you with the full text of this article because we are not able to identify you as a subscriber. If you have a personal subscription to this journal, then please login. If you are already logged in, then you may need to update your profile to register your subscription. Read more about accessing full-text

Abstract

We consider spectral properties of sparse sample covariance matrices, which includes biadjacency matrices of the bipartite Erdős–Rényi graph model. We prove a local law for the eigenvalue density up to the upper spectral edge. Under a suitable condition on the sparsity, we also prove that the limiting distribution of the rescaled, shifted extremal eigenvalues is given by the GOE Tracy–Widom law with an explicit formula on the deterministic shift of the spectral edge. For the biadjacency matrix of an Erdős–Rényi graph with two vertex sets of comparable sizes $M$ and $N$, this establishes Tracy–Widom fluctuations of the second largest eigenvalue when the connection probability $p$ is much larger than $N^{-2/3}$ with a deterministic shift of order $(Np)^{-1}$.

Article information

Source
Ann. Appl. Probab., Volume 29, Number 5 (2019), 3006-3036.

Dates
Received: August 2018
Revised: December 2018
First available in Project Euclid: 18 October 2019

Permanent link to this document
https://projecteuclid.org/euclid.aoap/1571385628

Digital Object Identifier
doi:10.1214/19-AAP1472

Mathematical Reviews number (MathSciNet)
MR4019881

Subjects
Primary: 60B20: Random matrices (probabilistic aspects; for algebraic aspects see 15B52) 62H10: Distribution of statistics

Keywords
High-dimensional sample covariance matrices local law Tracy–Widom distribution

Citation

Hwang, Jong Yun; Lee, Ji Oon; Schnelli, Kevin. Local law and Tracy–Widom limit for sparse sample covariance matrices. Ann. Appl. Probab. 29 (2019), no. 5, 3006--3036. doi:10.1214/19-AAP1472. https://projecteuclid.org/euclid.aoap/1571385628


Export citation

References

  • [1] Akhiezer, N. I. (1965). The Classical Moment Problem and Some Related Questions in Analysis. Hafner, New York. Translated by N. Kemmer.
  • [2] Anderson, G. W. (2013). Convergence of the largest singular value of a polynomial in independent Wigner matrices. Ann. Probab. 41 2103–2181.
  • [3] Bao, Z., Pan, G. and Zhou, W. (2015). Universality for the largest eigenvalue of sample covariance matrices with general population. Ann. Statist. 43 382–421.
  • [4] Bianchi, P., Debbah, M., Maida, M. and Najim, J. (2011). Performance of statistical tests for single-source detection using random matrix theory. IEEE Trans. Inform. Theory 57 2400–2419.
  • [5] Bickel, P. J. and Sarkar, P. (2016). Hypothesis testing for automated community detection in networks. J. R. Stat. Soc. Ser. B. Stat. Methodol. 78 253–273.
  • [6] Bourgade, P., Erdös, L. and Yau, H.-T. (2014). Edge universality of beta ensembles. Comm. Math. Phys. 332 261–353.
  • [7] Bourgade, P., Huang, J. and Yau, H.-T. (2017). Eigenvector statistics of sparse random matrices. Electron. J. Probab. 22 Paper No. 64.
  • [8] Choi, Y., Taylor, J. and Tibshirani, R. (2017). Selecting the number of principal components: Estimation of the true rank of a noisy matrix. Ann. Statist. 45 2590–2617.
  • [9] Ding, X. and Yang, F. (2018). A necessary and sufficient condition for edge universality at the largest singular values of covariance matrices. Ann. Appl. Probab. 28 1679–1738.
  • [10] El Karoui, N. (2007). Tracy–Widom limit for the largest eigenvalue of a large class of complex sample covariance matrices. Ann. Probab. 35 663–714.
  • [11] Erdős, L., Knowles, A., Yau, H.-T. and Yin, J. (2012). Spectral statistics of Erdős–Rényi graphs II: Eigenvalue spacing and the extreme eigenvalues. Comm. Math. Phys. 314 587–640.
  • [12] Erdős, L., Knowles, A., Yau, H.-T. and Yin, J. (2013). Spectral statistics of Erdős–Rényi graphs I: Local semicircle law. Ann. Probab. 41 2279–2375.
  • [13] Erdős, L., Yau, H.-T. and Yin, J. (2012). Rigidity of eigenvalues of generalized Wigner matrices. Adv. Math. 229 1435–1515.
  • [14] Feldman, V., Perkins, W. and Vempala, S. (2015). Subsampled power iteration: A unified algorithm for block models and planted CSP’s. In Advances in Neural Information Processing Systems 2836–2844.
  • [15] Florescu, L. and Perkins, W. (2016). Spectral thresholds in the bipartite stochastic block model. In Conference on Learning Theory 943–959.
  • [16] Girko, V. L. (1984). The circular law. Teor. Veroyatn. Primen. 29 669–679.
  • [17] Huang, J., Landon, B. and Yau, H.-T. Transition from Tracy–Widom to Gaussian fluctuations of extremal eigenvalues of sparse Erdős–Rényi graphs. Available at arXiv:1712.03936.
  • [18] Hwang, J. Y., Lee, J. O. and Schnelli, K. (2019). Supplement to “Local law and Tracy–Widom limit for sparse sample covariance matrices.” DOI:10.1214/19-AAP1472SUPP.
  • [19] Johansson, K. (2000). Shape fluctuations and random matrices. Comm. Math. Phys. 209 437–476.
  • [20] Johnstone, I. M. (2001). On the distribution of the largest eigenvalue in principal components analysis. Ann. Statist. 29 295–327.
  • [21] Knowles, A. and Yin, J. (2017). Anisotropic local laws for random matrices. Probab. Theory Related Fields 169 257–352.
  • [22] Kritchman, S. and Nadler, B. (2008). Determining the number of components in a factor model from limited noisy data. Chemom. Intell. Lab. Syst. 94 19–32.
  • [23] Lee, J. O. and Schnelli, K. (2015). Edge universality for deformed Wigner matrices. Rev. Math. Phys. 27 1550018.
  • [24] Lee, J. O. and Schnelli, K. (2016). Tracy–Widom distribution for the largest eigenvalue of real sample covariance matrices with general population. Ann. Appl. Probab. 26 3786–3839.
  • [25] Lee, J. O. and Schnelli, K. (2018). Local law and Tracy–Widom limit for sparse random matrices. Probab. Theory Related Fields 171 543–616.
  • [26] Lei, J. (2016). A goodness-of-fit test for stochastic block models. Ann. Statist. 44 401–424.
  • [27] Lytova, A. and Pastur, L. (2009). Central limit theorem for linear eigenvalue statistics of random matrices with independent entries. Ann. Probab. 37 1778–1840.
  • [28] Marčenko, V. A. and Pastur, L. A. (1967). Distribution of eigenvalues in certain sets of random matrices. Mat. Sb. 72 507–536.
  • [29] Nadakuditi, R. R. and Edelman, A. (2008). Sample eigenvalue based detection of high-dimensional signals in white noise using relatively few samples. IEEE Trans. Signal Process. 56 2625–2638.
  • [30] Onatski, A. (2009). Testing hypotheses about the numbers of factors in large factor models. Econometrica 77 1447–1479.
  • [31] Pillai, N. S. and Yin, J. (2014). Universality of covariance matrices. Ann. Appl. Probab. 24 935–1001.
  • [32] Quandt, R. E. (1972). A new approach to estimating switching regressions. J. Amer. Statist. Assoc. 67 306–310.
  • [33] Stein, C. M. (1981). Estimation of the mean of a multivariate normal distribution. Ann. Statist. 9 1135–1151.
  • [34] Timmermann, A. (2000). Moments of Markov switching models. J. Econometrics 96 75–111.

Supplemental materials

  • Supplement: Proofs of some lemmas. In the Supplementary Material [18], we will provide the proofs of Lemmas 4.1, 4.2, 4.4 and 5.4.