Propagation of outliers in multivariate data

Abstract

We investigate the performance of robust estimates of multivariate location under nonstandard data contamination models such as componentwise outliers (i.e., contamination in each variable is independent from the other variables). This model brings up a possible new source of statistical error that we call "propagation of outliers." This source of error is unusual in the sense that it is generated by the data processing itself and takes place after the data has been collected. We define and derive the influence function of robust multivariate location estimates under flexible contamination models and use it to investigate the effect of propagation of outliers. Furthermore, we show that standard high-breakdown affine equivariant estimators propagate outliers and therefore show poor breakdown behavior under componentwise contamination when the dimension $d$ is high.

Published in at http://dx.doi.org/10.1214/07-AOS588 the Annals of Statistics (http://www.imstat.org/aos/) by the Institute of Mathematical Statistics (http://www.imstat.org)

Cited by in corpus (35)

• Anomaly Detection by Robust Statistics
• Multivariate outlier detection based on a robust Mahalanobis distance with shrinkage estimators
• Detecting deviating data cells
• Robust estimation of precision matrices under cellwise contamination
• Fast robust correlation for high-dimensional data
• MacroPCA: An all-in-one PCA method allowing for missing values as well as cellwise and rowwise outliers
• Robust VIF regression with application to variable selection in large data sets
• The shooting S-estimator for robust regression
• The Cellwise Minimum Covariance Determinant Estimator
• Robust Identification of Target Genes and Outliers in Triple-negative Breast Cancer Data
• Multivariate outlier explanations using Shapley values and Mahalanobis distances
• Challenges of cellwise outliers
• Robust regression estimation and inference in the presence of cellwise and casewise contamination
• Robust Multivariate Functional Control Chart
• Handling cellwise outliers by sparse regression and robust covariance
• Simultaneous Feature Selection and Outlier Detection with Optimality Guarantees
• Resistant estimates for high dimensional and functional data based on random projections
• Robust Multivariate Estimation Based On Statistical Depth Filters
• Multivariate Singular Spectrum Analysis by Robust Diagonalwise Low-Rank Approximation
• Analyzing cellwise weighted data
• Robust covariance estimation and explainable outlier detection for matrix-valued data
• A Robust Seemingly Unrelated Regressions For Row-Wise And Cell-Wise Contamination
• Cellwise outlier detection in heterogeneous populations
• Robust Variable Selection under Cellwise Contamination
• Anomaly Detection via Graphical Lasso
• Doubly Robust Feature Selection with Mean and Variance Outlier Detection and Oracle Properties
• The column measure and Gradient-Free Gradient Boosting
• Wavelet Variance for Random Fields: an M-Estimation Framework
• Robust high-dimensional precision matrix estimation
• Robust variable screening for regression using factor profiling
• Robust estimation of multivariate location and scatter in the presence of cellwise and casewise contamination
• Robust Orthogonal Complement Principal Component Analysis
• Composite Robust Estimators for Linear Mixed Models
• Casewise and Cellwise Robust Multilinear Principal Component Analysis
• Robust Principal Components by Casewise and Cellwise Weighting