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.

A High Dimensional Nonparametric Test for Proportional Covariance Matrices

A hypothesis test for independence of sets of variates in high dimensions

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

Tests for High-Dimensional Covariance Matrices Using Random Matrix Projection

Two Sample Tests for High Dimensional Covariance Matrices

HYPOTHESIS TESTING ON LINEAR STRUCTURES OF HIGH DIMENSIONAL COVARIANCE MATRIX.

Tests for Covariance Structures with High-Dimensional Repeated Measurements

Use of Random Integration to Test Equality of High Dimensional Covariance Matrices

Many-sample tests for the equality and the proportionality hypotheses between large covariance matrices

A Unified Framework for Testing High Dimensional Parameters: A Data-Adaptive Approach

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

A Unified Framework for Testing High Dimensional Parameters: A Data-Adaptive Approach.

Two sample test for covariance matrices in ultra-high dimension

Double verification for two‐sample covariance matrices test

Testing Super-Diagonal Structure in High Dimensional Covariance Matrices

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

Detecting Change Points of Covariance Matrices in High Dimensions

Testing the Diagonality of A Large Covariance Matrix in A Regression Setting

HIGH-DIMENSIONAL TWO-SAMPLE COVARIANCE MATRIX TESTING VIA SUPER-DIAGONALS

Testing and Confidence Intervals for High Dimensional Proportional Hazards Model