On the robustness to adversarial corruption and to heavy-tailed data of the Stahel–Donoho median of means

DOI10.1093/IMAIAI/IAAC026MaRDI QIDQ6039760zbMATH OpenOpenAlexFDO

Authors Jules Depersin, Guillaume Lecué

Publication date 23 May 2023

Published in Information and Inference: A Journal of the IMA (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/2101.09117

zbMATH Keywords

\(d\)-dimensional medianadversarial contaminationdepthheavy-tailed datamedian of meansrobustness

Mathematics Subject Classification ID

Robustness and adaptive procedures (parametric inference) (62F35) General topics in the theory of data (68P01)

Abstract: We consider median of means (MOM) versions of the Stahel-Donoho outlyingness (SDO) [stahel 1981, donoho 1982] and of Median Absolute Deviation (MAD) functions to construct subgaussian estimators of a mean vector under adversarial contamination and heavy-tailed data. We develop a single analysis of the MOM version of the SDO which covers all cases ranging from the Gaussian case to the L2 case. It is based on isomorphic and almost isometric properties of the MOM versions of SDO and MAD. This analysis also covers cases where the mean does not even exist but a location parameter does; in those cases we still recover the same subgaussian rates and the same price for adversarial contamination even though there is not even a first moment. These properties are achieved by the classical SDO median and are therefore the first non-asymptotic statistical bounds on the Stahel-Donoho median complementing the $sqrtn$-consistency [maronna 1995] and asymptotic normality [Zuo, Cui, He, 2004] of the Stahel-Donoho estimators. We also show that the MOM version of MAD can be used to construct an estimator of the covariance matrix under only a L2-moment assumption or of a scale parameter if a second moment does not exist.

Recommendations

• Robust sub-Gaussian estimation of a mean vector in nearly linear time
• Robust machine learning by median-of-means: theory and practice
• Mean estimation and regression under heavy-tailed distributions: A survey
• Learning from MOM's principles: Le Cam's approach
• The Behavior of the Stahel-Donoho Robust Multivariate Estimator

Cited in

(6)

• Non-asymptotic analysis and inference for an outlyingness induced winsorized mean
• The curse of overparametrization in adversarial training: precise analysis of robust generalization for random features regression
• Exponential concentration for geometric-median-of-means in non-positive curvature spaces
• Differentially private projection-depth-based medians
• Statistically optimal robust mean and covariance estimation for anisotropic gaussians
• On the breakdown point of transport-based quantiles

This page was built for publication: On the robustness to adversarial corruption and to heavy-tailed data of the Stahel–Donoho median of means

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6039760)