Robust Sparse Covariance Estimation by Thresholding Tyler's M-Estimator
arXiv:1706.08020 · doi:10.1214/18-AOS1793
Abstract
Estimating a high-dimensional sparse covariance matrix from a limited number of samples is a fundamental problem in contemporary data analysis. Most proposals to date, however, are not robust to outliers or heavy tails. Towards bridging this gap, in this work we consider estimating a sparse shape matrix from samples following a possibly heavy tailed elliptical distribution. We propose estimators based on thresholding either Tyler's M-estimator or its regularized variant. We derive bounds on the difference in spectral norm between our estimators and the shape matrix in the joint limit as the dimension and sample size tend to infinity with . These bounds are minimax rate-optimal. Results on simulated data support our theoretical analysis.
References in corpus (15)
- Covariance regularization by thresholding
- Sparsistency and rates of convergence in large covariance matrix estimation
- Robust Shrinkage Estimation of High-dimensional Covariance Matrices
- Operator norm consistent estimation of large-dimensional sparse covariance matrices
- Optimal rates of convergence for sparse covariance matrix estimation
- Generalized robust shrinkage estimator and its application to STAP detection problem
- Regularized -estimators of scatter matrix
- Robust Estimation of Structured Covariance Matrix for Heavy-Tailed Elliptical Distributions
- Regularized Tyler's Scatter Estimator: Existence, Uniqueness, and Algorithms
- Robust Estimates of Covariance Matrices in the Large Dimensional Regime
- Tyler's Covariance Matrix Estimator in Elliptical Models with Convex Structure
- A cautionary note on robust covariance plug-in methods
- M-Functionals of Multivariate Scatter
- Large Dimensional Analysis of Robust M-Estimators of Covariance with Outliers
- New Algorithms for -Estimation of Multivariate Scatter and Location