Asymptotics of Empirical Eigen-structure for Ultra-high Dimensional Spiked Covariance Model

Abstract

We derive the asymptotic distributions of the spiked eigenvalues and eigenvectors under a generalized and unified asymptotic regime, which takes into account the spike magnitude of leading eigenvalues, sample size, and dimensionality. This new regime allows high dimensionality and diverging eigenvalue spikes and provides new insights into the roles the leading eigenvalues, sample size, and dimensionality play in principal component analysis. The results are proven by a technical device, which swaps the role of rows and columns and converts the high-dimensional problems into low-dimensional ones. Our results are a natural extension of those in Paul (2007) to more general setting with new insights and solve the rates of convergence problems in Shen et al. (2013). They also reveal the biases of the estimation of leading eigenvalues and eigenvectors by using principal component analysis, and lead to a new covariance estimator for the approximate factor model, called shrinkage principal orthogonal complement thresholding (S-POET), that corrects the biases. Our results are successfully applied to outstanding problems in estimation of risks of large portfolios and false discovery proportions for dependent test statistics and are illustrated by simulation studies.

50 pages, 6 figures

References in corpus (3)

• Convergence and prediction of principal component scores in high-dimensional settings
• Surprising Asymptotic Conical Structure in Critical Sample Eigen-Directions
• Concentration Inequalities and Moment Bounds for Sample Covariance Operators

Cited by in corpus (12)

• Large Covariance Estimation through Elliptical Factor Models
• An ${\ell }_{\infty }$ Eigenvector Perturbation Bound and Its Application to Robust Covariance Estimation
• De-biased sparse PCA: Inference and testing for eigenstructure of large covariance matrices
• A Model-free Variable Screening Method Based on Leverage Score
• Asymmetry Helps: Eigenvalue and Eigenvector Analyses of Asymmetrically Perturbed Low-Rank Matrices
• Distributed Estimation for Principal Component Analysis: an Enlarged Eigenspace Analysis
• Robust Covariance Estimation for Approximate Factor Models
• Heterogeneity Adjustment with Applications to Graphical Model Inference
• The limits of the sample spiked eigenvalues for a high-dimensional generalized Fisher matrix and its applications
• Non-Sparse PCA in High Dimensions via Cone Projected Power Iteration
• Bayesian Factor-adjusted Sparse Regression
• Estimation of the number of spikes using a generalized spike population model and application to RNA-seq data