Abstract:Testing large covariance matrices is of fundamental importance in statistical analysis with high-dimensional data. In the past decade, three types of test statistics have been studied in the literature: quadratic form statistics, maximum form statistics, and their weighted combination. It is known that quadratic form statistics would suffer from low power against sparse alternatives and maximum form statistics would suffer from low power against dense alternatives. The weighted combination methods were introduced to enhance the power of quadratic form statistics or maximum form statistics when the weights are appropriately chosen. In this article, we provide a new perspective to exploit the full potential of quadratic form statistics and maximum form statistics for testing high-dimensional covariance matrices. We propose a scale-invariant power-enhanced test based on Fisher's method to combine the p -values of quadratic form statistics and maximum form statistics. After carefully studying the asymptotic joint distribution of quadratic form statistics and maximum form statistics, we first prove that the proposed combination method retains the correct asymptotic size under the Gaussian assumption, and we also derive a new Lyapunov-type bound for the joint distribution and prove the correct asymptotic size of the proposed method without requiring the Gaussian assumption. Moreover, we show that the proposed method boosts the asymptotic power against more general alternatives. Finally, we demonstrate the finite-sample performance in simulation studies and a real application. Supplementary materials for this article are available online.

Test for Bandedness of High-Dimensional Covariance Matrices and Bandwidth Estimation

Band Width Selection for High Dimensional Covariance Matrix Estimation

More Powerful Tests for Sparse High-Dimensional Covariances Matrices

Convex Banding of the Covariance Matrix

On the theoretic and practical merits of the banding estimator for large covariance matrices

Multi-level Thresholding Test for High Dimensional Covariance Matrices

Testing Super-Diagonal Structure in High Dimensional Covariance Matrices

Two Sample Tests for High Dimensional Covariance Matrices

Convergence Rates of Spectral Distributions of Large Sample Covariance Matrices

EIGENVALUE DISTRIBUTIONS OF VARIANCE COMPONENTS ESTIMATORS IN HIGH-DIMENSIONAL RANDOM EFFECTS MODELS

Asymptotic Performance Analysis of Bearing Estimate for Spatially Distributed Source with Finite Bandwidth

A Note on the Convergence Rate of the Spectral Distributions of Large Random Matrices

On testing the equality of high dimensional mean vectors with unequal covariance matrices

Testing High-dimensional Covariance Matrices under the Elliptical Distribution and Beyond

Identity Tests for High Dimensional Data Using Rmt

A Note on the Convergence Rate of the Spectral Distributions of Large Sample Covariance Matrices

Regularized estimation of large covariance matrices

Two sample test for covariance matrices in ultra-high dimension

Fisher's Combined Probability Test for High-Dimensional Covariance Matrices

Testing and Signal Identification for Two-Sample High-Dimensional Covariances Via Multi-Level Thresholding

On the inference about the spectral distribution of high-dimensional covariance matrix based on high-frequency noisy observations