Abstract:We investigate the quantitative performance of affine-equivariant estimators for robust mean estimation. As a natural stability requirement, the construction of such affine-equivariant estimators has been extensively studied in the statistics literature. We quantitatively evaluate these estimators under two outlier models which have been the subject of much recent work: the heavy-tailed and adversarial corruption settings. We establish lower bounds which show that affine-equivariance induces a strict degradation in recovery error with quantitative rates degrading by a factor of $\sqrt{d}$ in both settings. We find that classical estimators such as the Tukey median (Tukey '75) and Stahel-Donoho estimator (Stahel '81 and Donoho '82) are either quantitatively sub-optimal even within the class of affine-equivariant estimators or lack any quantitative guarantees. On the other hand, recent estimators with strong quantitative guarantees are not affine-equivariant or require additional distributional assumptions to achieve it. We remedy this by constructing a new affine-equivariant estimator which nearly matches our lower bound. Our estimator is based on a novel notion of a high-dimensional median which may be of independent interest. Notably, our results are applicable more broadly to any estimator whose performance is evaluated in the Mahalanobis norm which, for affine-equivariant estimators, corresponds to an evaluation in Euclidean norm on isotropic distributions.

A new family of robust quantile-regression-based mean estimators using Sarndal approach

Compromised imputation based mean estimators using robust quantile regression

Improved Ratio Estimators Using Some Robust Measures

Enhanced Estimation of Population Mean in the Presence of Auxiliary Information

Optimized estimation for population mean using conventional and non-conventional measures under the joint influence of measurement error and non-response

On Improved Estimation of Population Mean Using Known Coefficient of Skewness of an Auxiliary Variable

Dual-type quantile regression approach for mean estimation incorporating sampled and non-sampled data

Improved Estimation of Population Mean Through Known Conventional and Non-Conventional Measures of Auxiliary Variable

Improved Ratio Type Estimators of Population Mean Based on Median of a Study Variable and an Auxiliary Variable

A redescending M-estimator approach for outlier-resilient modeling

Three-fold utilization of supplementary information for mean estimation under median ranked set sampling scheme

Enhancing performance in the presence of outliers with redescending M-estimators

Small Area Quantile Estimation

Use Of Nonconventional Dispersion Measures To Improve The Efficiency Of Ratio-Type Estimators Of Variance In The Presence Of Outliers

Bootstrap-quantile ridge estimator for linear regression with applications

Double Exponential Ratio Estimator of a Finite Population Variance under Extreme Values in Simple Random Sampling

Statistical Barriers to Affine-equivariant Estimation

Enhanced direct and synthetic estimators for domain mean with simulation and applications

On exploring the generalized mixture estimators under simple random sampling and application in health and finance sector

Robust maximum Lq-likelihood estimation of joint mean–covariance models for longitudinal data

Outlier-robust Mean Estimation near the Breakdown Point via Sum-of-Squares