Robust sparse covariance estimation by thresholding Tyler's M-estimator

DOI10.1214/18-AOS1793zbMATH Open1439.62123arXiv1706.08020MaRDI QIDQ2176609FDOQ2176609

Authors: John Goes, Gilad Lerman, Boaz Nadler 

Publication date: 5 May 2020

Published in: The Annals of Statistics (Search for Journal in Brave)

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 $n$ 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 $p$ and sample size $n$ tend to infinity with $p/nogamma>0$. These bounds are minimax rate-optimal. Results on simulated data support our theoretical analysis.

Full work available at URL: https://arxiv.org/abs/1706.08020

Recommendations

• Structured robust covariance estimation
• Trimmed estimators for large dimensional sparse covariance matrices
• High-dimensional covariance estimation under the presence of outliers
• Minimax optimal estimation of high-dimensional sparse covariance matrices with missing data
• Robust covariance and scatter matrix estimation under Huber's contamination model

zbMATH Keywords

covariance matrix estimationsparsityelliptical distributionthresholdingspectral normTyler's M-estimator

Mathematics Subject Classification ID

Asymptotic properties of nonparametric inference (62G20) Nonparametric robustness (62G35) Estimation in multivariate analysis (62H12) Inference from stochastic processes and spectral analysis (62M15)

Cites Work

• Title not available (Why is that?)
• Title not available (Why is that?)
• Covariance regularization by thresholding
• Sparsistency and rates of convergence in large covariance matrix estimation
• A distribution-free M-estimator of multivariate scatter
• Robust m-estimators of multivariate location and scatter
• Robust Statistics
• Hanson-Wright inequality and sub-Gaussian concentration
• On the theory of elliptically contoured distributions
• Local operator theory, random matrices and Banach spaces.
• Title not available (Why is that?)
• Title not available (Why is that?)
• Operator norm consistent estimation of large-dimensional sparse covariance matrices
• Redescending \(M\)-estimates of multivariate location and scatter
• \(M\)-functionals of multivariate scatter
• Adaptive Thresholding for Sparse Covariance Matrix Estimation
• Statistical analysis for the angular central Gaussian distribution on the sphere
• Generalized Thresholding of Large Covariance Matrices
• Optimal rates of convergence for sparse covariance matrix estimation
• New algorithms for \(M\)-estimation of multivariate scatter and location
• A generalization of Tyler's M-estimators to the case of incomplete data
• Title not available (Why is that?)
• A cautionary note on robust covariance plug-in methods
• On Tyler's \(M\)-functional of scatter in high dimension
• The random matrix regime of Maronna's M-estimator with elliptically distributed samples
• Marčenko-Pastur law for Tyler's M-estimator
• Robust Estimates of Covariance Matrices in the Large Dimensional Regime
• Robust Shrinkage Estimation of High-Dimensional Covariance Matrices
• Unified Framework to Regularized Covariance Estimation in Scaled Gaussian Models
• Tyler's Covariance Matrix Estimator in Elliptical Models With Convex Structure
• Large Dimensional Analysis of Robust M-Estimators of Covariance With Outliers
• Structured Robust Covariance Estimation
• Second order statistics of robust estimators of scatter. Application to GLRT detection for elliptical signals
• Concentration of the spectral measure for large matrices
• Large dimensional analysis and optimization of robust shrinkage covariance matrix estimators
• Symmetrised M-estimators of multivariate scatter
• Robust and efficient estimation of multivariate scatter and location
• Estimating structured high-dimensional covariance and precision matrices: optimal rates and adaptive estimation
• Robust covariance and scatter matrix estimation under Huber's contamination model
• Generalized Robust Shrinkage Estimator and Its Application to STAP Detection Problem
• Robust estimation of high-dimensional covariance and precision matrices
• Regularized <formula formulatype="inline"><tex Notation="TeX">$M$</tex> </formula>-Estimators of Scatter Matrix
• The sample covariance is not efficient for elliptical distributions
• Robust Estimation of Structured Covariance Matrix for Heavy-Tailed Elliptical Distributions
• Robust sparse covariance estimation by thresholding Tyler's M-estimator
• Regularized Tyler's Scatter Estimator: Existence, Uniqueness, and Algorithms

Cited In (9)

• Robust semiparametric modeling of mean and covariance in longitudinal data
• A review of Tyler's shape matrix and its extensions
• Rejoinder to “A Tuning-Free Robust and Efficient Approach to High-Dimensional Regression”
• Tyler's and Maronna's M-estimators: non-asymptotic concentration results
• Covariance Model with General Linear Structure and Divergent Parameters
• Robust Covariance Matrix Estimation for High-Dimensional Compositional Data with Application to Sales Data Analysis
• Robust Shape Matrix Estimation for High-Dimensional Compositional Data with Application to Microbial Inter-Taxa Analysis
• An overview of heavy-tail extensions of multivariate Gaussian distribution and their relations
• Robust sparse covariance estimation by thresholding Tyler's M-estimator

This page was built for publication: Robust sparse covariance estimation by thresholding Tyler's M-estimator

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2176609)