Robust Shrinkage Estimation of High-dimensional Covariance Matrices

Abstract

We address high dimensional covariance estimation for elliptical distributed samples, which are also known as spherically invariant random vectors (SIRV) or compound-Gaussian processes. Specifically we consider shrinkage methods that are suitable for high dimensional problems with a small number of samples (large $p$ small $n$). We start from a classical robust covariance estimator [Tyler(1987)], which is distribution-free within the family of elliptical distribution but inapplicable when $n<p$. Using a shrinkage coefficient, we regularize Tyler's fixed point iterations. We prove that, for all $n$ and $p$, the proposed fixed point iterations converge to a unique limit regardless of the initial condition. Next, we propose a simple, closed-form and data dependent choice for the shrinkage coefficient, which is based on a minimum mean squared error framework. Simulations demonstrate that the proposed method achieves low estimation error and is robust to heavy-tailed samples. Finally, as a real world application we demonstrate the performance of the proposed technique in the context of activity/intrusion detection using a wireless sensor network.

References in corpus (2)

• Regularized estimation of large covariance matrices
• Shrinkage Algorithms for MMSE Covariance Estimation

Cited by in corpus (48)

• Conic geometric optimisation on the manifold of positive definite matrices
• Generalized robust shrinkage estimator and its application to STAP detection problem
• Regularized $M$-estimators of scatter matrix
• Covariance Estimation in High Dimensions via Kronecker Product Expansions
• Multivariate Generalized Gaussian Distribution: Convexity and Graphical Models
• Multivariate outlier detection based on a robust Mahalanobis distance with shrinkage estimators
• Machine Learning for Wireless Link Quality Estimation: A Survey
• Regularized Tyler's Scatter Estimator: Existence, Uniqueness, and Algorithms
• Radar Adaptive Detection Architectures for Heterogeneous Environments
• A Robust Statistics Approach to Minimum Variance Portfolio Optimization
• Tyler's Covariance Matrix Estimator in Elliptical Models with Convex Structure
• Performance Analysis of Tyler's Covariance Estimator
• Robust Sparse Covariance Estimation by Thresholding Tyler's M-Estimator
• Learning Latent Variable Gaussian Graphical Models
• Large Dimensional Analysis of Robust M-Estimators of Covariance with Outliers
• Space-Time Adaptive Detection at Low Sample Support
• New insights into the statistical properties of $M$-estimators
• Regularized Covariance Estimation for Polarization Radar Detection in Compound Gaussian Sea Clutter
• Optimal Design of the Adaptive Normalized Matched Filter Detector
• Bounds for maximum likelihood regular and non-regular DoA estimation in $K$-distributed noise
• Learning to Fairly Classify the Quality of Wireless Links
• Large-dimensional behavior of regularized Maronna's M-estimators of covariance matrices
• Convergence and Fluctuations of Regularized Tyler Estimators
• R-NL: Covariance Matrix Estimation for Elliptical Distributions based on Nonlinear Shrinkage
• Convex programming approach to robust estimation of a multivariate Gaussian model
• Expertise and Task Pressure in fNIRS-based brain Connectomes
• Physics-Informed Machine Learning with Conditional Karhunen-Loève Expansions
• A Stochastic Covariance Shrinkage Approach to Particle Rejuvenation in the Ensemble Transform Particle Filter
• Group Symmetric Robust Covariance Estimation
• Regularized Block Toeplitz Covariance Matrix Estimation via Kronecker Product Expansions
• Robust self-tuning semiparametric PCA for contaminated elliptical distribution
• Marchenko-Pastur Law for Tyler's M-estimator
• Tyler's and Maronna's M-estimators: Non-Asymptotic Concentration Results
• Linear pooling of sample covariance matrices
• Cross-Validated Tuning of Shrinkage Factors for MVDR Beamforming Based on Regularized Covariance Matrix Estimation
• Principal Regression for High Dimensional Covariance Matrices
• Joint Covariance Estimation with Mutual Linear Structure
• Structured Channel Covariance Estimation from Limited Samples for Large Antenna Arrays
• Constrained Bayesian ICA for Brain Connectome Inference
• Eigengap Sparsity for Covariance Parsimony
• Spectrum Sensing with Small-Sized Datasets in Cognitive Radio: Algorithms and Analysis
• Study of Efficient Robust Adaptive Beamforming Algorithms Based on Shrinkage Techniques
• Gaussian and Robust Kronecker Product Covariance Estimation: Existence and Uniqueness
• High-Dimensional Covariance Shrinkage for Signal Detection
• EMFlow: Data Imputation in Latent Space via EM and Deep Flow Models
• Covariance Estimation in Elliptical Models with Convex Structure
• Ensemble Kalman Filter Implementations Based on Covariance Matrix Estimation
• Adaptive Shrinkage Estimation for Streaming Graphs