Non-asymptotic robustness analysis of regression depth median

DOI10.1016/J.JMVA.2023.105247MaRDI QIDQ6183697zbMATH OpenFDO

Authors Yijun Zuo

Publication date 4 January 2024

Published in Journal of Multivariate Analysis (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/2009.00646

zbMATH Keywords

finite-sample breakdown pointregression depthregression medianmaximum regression depth estimatornon-asymptotic robustness analysis

Mathematics Subject Classification ID

Nonparametric regression and quantile regression (62G08) Nonparametric robustness (62G35) Robustness and adaptive procedures (parametric inference) (62F35) Multivariate analysis (62Hxx)

Abstract: The maximum depth estimator (aka depth median) ($$) induced from regression depth (RD) of Rousseeuw and Hubert (1999) (RH99) is one of the most prevailing estimators in regression. It possesses outstanding robustness similar to the univariate location counterpart. Indeed, $$ can, asymptotically, resist up to $33\mathrm{\%}$ contamination without breakdown, in contrast to the $0\mathrm{\%}$ for the traditional (least squares and least absolute deviations) estimators (see Van Aelst and Rousseeuw, 2000) (VAR00)). The results from VAR00 are pioneering, yet they are limited to regression-symmetric populations (with a strictly positive density) and the $epsilon$-contamination and maximum-bias model. With a fixed finite-sample size practice, the most prevailing measure of robustness for estimators is the finite-sample breakdown point (FSBP) (Donoho and Huber (1983)). Despite many attempts made in the literature, only sporadic partial results on FSBP for $$ were obtained whereas an exact FSBP for $$ remained open in the last twenty-plus years. Furthermore, is the asymptotic breakdown value $1/3$ (the limit of an increasing sequence of finite-sample breakdown values) relevant in the finite-sample practice? (Or what is the difference between the finite-sample and the limit breakdown values?). Such discussions are yet to be given in the literature. This article addresses the above issues, revealing an intrinsic connection between the regression depth of $$ and the newly obtained exact FSBP. It justifies the employment of $$ as a robust alternative to the traditional estimators and demonstrates the necessity and the merit of using the FSBP in finite-sample real practice.

Cites work

• scientific article; zbMATH DE number 3829050 (Why is no real title available?)
• scientific article; zbMATH DE number 3954047 (Why is no real title available?)
• scientific article; zbMATH DE number 3541764 (Why is no real title available?)
• scientific article; zbMATH DE number 194744 (Why is no real title available?)
• scientific article; zbMATH DE number 3335619 (Why is no real title available?)
• A General Qualitative Definition of Robustness
• Aspects of robust linear regression
• Asymptotic behaviour of S-estimates of multivariate location parameters and dispersion matrices
• Breakdown points for designed experiments
• Breakdown points of affine equivariant estimators of multivariate location and covariance matrices
• Breakdown properties of location estimates based on halfspace depth and projected outlyingness
• Characterizing angular symmetry and regression symmetry.
• Computing halfspace depth and regression depth
• Finite sample breakdown of M- and P-estimators
• Finite sample breakdown point of Tukey's halfspace median
• Finite sample breakdown points of projection based multivariate location and scatter statistics
• Highly efficient estimators of multivariate location with high breakdown point
• Least Median of Squares Regression
• On depth and deep points: A calculus.
• On general notions of depth for regression
• On multivariate monotonic measures of location with high breakdown point
• Regression Depth
• Robust Statistics
• Robust Statistics
• Robustness of deepest regression
• Robustness of the deepest projection regression functional
• Robustness of the half-space median
• Some quantitative relationships between two types of finite sample breakdown point
• The Breakdown Point of Simultaneous General M Estimates of Regression and Scale
• The asymptotics of S-estimators in the linear regression model
• The breakdown point -- examples and counterexamples
• The deepest regression method
• The influence function and maximum bias of Tukey's median

This page was built for publication: Non-asymptotic robustness analysis of regression depth median

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6183697)