A homogeneity test of large dimensional covariance matrices under non-normality

Ahmad, M. Rauf

Kybernetika, Tome 54 (2018) no. 5, pp. 908-920

Abstract (VO)

A test statistic for homogeneity of two or more covariance matrices is presented when the distributions may be non-normal and the dimension may exceed the sample size. Using the Frobenius norm of the difference of null and alternative hypotheses, the statistic is constructed as a linear combination of consistent, location-invariant, estimators of trace functions that constitute the norm. These estimators are defined as $U$-statistics and the corresponding theory is exploited to derive the normal limit of the statistic under a few mild assumptions as both sample size and dimension grow large. Simulations are used to assess the accuracy of the statistic.

Détail

DOI : 10.14736/kyb-2018-5-0908

Classification : 62H15
Keywords: high-dimensional inference; covariance testing; $U$-statistics; non-normality

Bibliographie

[1] Ahmad, M. R.: Location-invariant multi-sample $U$-tests for covariance matrices with large dimension. Scand. J. Stat. 44 (2017b), 500-523. | DOI | MR

[2] Ahmad, M. R.: Testing homogeneity of several covariance matrices and multi-sample sphericity for high-dimensional data under non-normality. Comm. Stat. Theory Methods 46 (2017a), 3738-3753. | DOI | MR

[3] Ahmad, M. R.: On testing sphericity and identity of a covariance matrix with large dimensions. Math. Meth. Stat. 25 (2016), 121-132. | DOI | MR

[4] Anderson, T.W.: An Introduction to Multivariate Statistical Analysis. Third edition. Wiley, NY 2003. | MR

[5] Aoshima, M., Yata, K.: Two-stage procedures for high-dimensional data. Seq. An. 30 (2011), 356-399. | DOI | MR

[6] Cai, T., Liu, W., Xia, Y.: Two-sample covariance matrix testing and support recovery in high-dimensional and sparse settings. J. Amer. Statist. Assoc. 108 (2013), 265-277. | DOI | MR

[7] Fujikoshi, Y., Ulyanov, V. V., Shimizu, R.: Multivariate statistics: High-dimensional and large-sample approximations. Wiley, NY 2010. | MR

[8] Hájek, J., Šidák, Z., Sen, P. K.: Theory of Rank Tests. Academic Press, SD 1999. | MR | Zbl

[9] Kim, T. Y., Luo, Z-M., Kim, C.: The central limit theorem for degenerate variable $U$-statistics under dependence. J. Nonparam. Stat. 23 (2011), 683-699. | DOI | MR

[10] Koroljuk, V. S., Borovskich, Y. V.: Theory of $U$-statistics. Kluwer Academic Press, Dordrecht 1994. | DOI | MR

[11] Lee, A. J.: $U$-statistics: Theory and Practice. CRC Press, Boca Raton 1990.

[12] Lehmann, E. L.: Elements of Large-sample Theory. Springer, NY 1999. | DOI | MR

[13] Li, J., Chen, S. X.: Two sample tests for high-dimensional covariance matrices. Ann. Stat. 40 (2012), 908-940. | DOI | MR

[14] Liu, B., Xu, L., Zheng, S., Tian, G-L.: A new test for the proportionality of two large-dimensional covariance matrices. J. Multiv. An. 131 (2014), 293-308. | DOI | MR

[15] Mikosch, T.: Weak invariance principles for weighted $U$-statistics. J. Theoret. Prob. 7 (1991), 147-173. | DOI | MR

[16] Mikosch, T.: A weak invariance principle for weighted $U$-statistics with varying kernels. J. Multiv. An. 47 (1993), 82-102. | DOI | MR

[17] Muirhead, R. J.: Aspects of Multivariate Statistical Theory. Wiley, NY 2005 | DOI | MR

[18] Pinheiro, A., Sen, K., Pinheiro, H. P.: Decomposibility of high-dimensional diversity measures: Quasi-$U$-statistics, martigales, and nonstandard asymptotics. J. Multiv. An. 100 (2009), 1645-1656. | DOI | MR

[19] Qiu, Y., Chen, S. X.: Test for bandedness of high-dimensional covariance matrices and bandwidth estimation. Ann. Stat. 40 (2012) 1285-1314. | DOI | MR

[20] Schott, J. R.: A test for the equality of covariance matrices when the dimension is large relative to the sample size. Computat. Statist. Data Analysis 51 (2007), 6535-6542. | DOI | MR

[21] Seber, G. A. F.: Multivariate Observations. Wiley, NY 2004. | DOI | MR

[22] Sen, P. K.: Robust statistical inference for high-dimensional data models with applications in genomics. Aust. J. Stat. 35 (2006), 197-214.

[23] Serfling, R. J.: Approximation Theorems of Mathematical Statistics. Wiley, Weinheim 1980. | DOI | MR

[24] Srivastava, M. S., Yanagihara, H.: Testing the equality of several covariance matrices with fewer observations than the dimension. J. Multiv. An. 101, 1319-1329. | DOI | MR

[25] Vaart, A. W. van der: Asymptotic Statistics. Cambridge University Press, 1998. | DOI | MR

[26] Zhong, P-S., Chen, S. X.: Tests for high-dimensional regression coefficients with factorial designs. J. Amer. Statist. Assoc. 106 (2011), 260-274. | DOI | MR