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.

Likelihood-based Tests on Linear Hypotheses of Large Dimensional Mean Vectors with Unequal Covariance Matrices

Likelihood-Based Tests on Moderate-High-Dimensional Mean Vectors with Unequal Covariance Matrices

An Adaptive Test for the Mean Vector in Large-P-small-n Problems

Testing Linear Hypothesis of High-Dimensional Means with Unequal Covariance Matrices

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

Hypothesis Testing for High-Dimensional Covariance Matrices.

HYPOTHESIS TESTING ON LINEAR STRUCTURES OF HIGH DIMENSIONAL COVARIANCE MATRIX.

A Scale-Invariant Test for Linear Hypothesis of Means in High Dimensions

Simulation-Based Hypothesis Testing of High Dimensional Means Under Covariance Heterogeneity

Modifications of Wald's Score Tests on Large Dimensional Covariance Matrices Structure

A Simultaneous Test of Mean Vector and Covariance Matrix in High-Dimensional Settings

A High Dimensional Nonparametric Test for Proportional Covariance Matrices

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

Test for high-dimensional linear hypothesis of mean vectors via random integration

Joint Test for Homogeneity of High-Dimensional Means and Covariance Matrices Using Maximum-Type Statistics

Likelihood ratio tests under model misspecification in high dimensions

High-dimensional general linear hypothesis tests via non-linear spectral shrinkage

General linear hypothesis testing of high-dimensional mean vectors with unequal covariance matrices based on random integration

Towards a theory for testing statistical hypothesis: Multivariate mean with nuisance covariance matrix

Testing Linear Hypotheses in High-Dimensional Regressions

Two Sample Tests for High Dimensional Covariance Matrices