Multidimensional trimming based on projection depth

DOI10.1214/009053606000000713zbMATH Open1106.62057arXivmath/0702655OpenAlexW1969901539MaRDI QIDQ869968FDOQ869968

Authors: Yijun Zuo 

Publication date: 12 March 2007

Published in: The Annals of Statistics (Search for Journal in Brave)

Abstract: As estimators of location parameters, univariate trimmed means are well known for their robustness and efficiency. They can serve as robust alternatives to the sample mean while possessing high efficiencies at normal as well as heavy-tailed models. This paper introduces multidimensional trimmed means based on projection depth induced regions. Robustness of these depth trimmed means is investigated in terms of the influence function and finite sample breakdown point. The influence function captures the local robustness whereas the breakdown point measures the global robustness of estimators. It is found that the projection depth trimmed means are highly robust locally as well as globally. Asymptotics of the depth trimmed means are investigated via those of the directional radius of the depth induced regions. The strong consistency, asymptotic representation and limiting distribution of the depth trimmed means are obtained. Relative to the mean and other leading competitors, the depth trimmed means are highly efficient at normal or symmetric models and overwhelmingly more efficient when these models are contaminated. Simulation studies confirm the validity of the asymptotic efficiency results at finite samples.

Full work available at URL: https://arxiv.org/abs/math/0702655

Recommendations

• Multivariate trimmed means based on the Tukey depth
• Influence function and maximum bias of projection depth based estimators.
• Asymptotic properties of location estimators based on projection depth
• Breakdown properties of location estimates based on halfspace depth and projected outlyingness
• On a robust and efficient maximum depth estimator

zbMATH Keywords

asymptoticsefficiencyrobustnessinfluence functionbreakdown pointdepth regionsdirectional radiusmultivariate trimmed means

Mathematics Subject Classification ID

Asymptotic properties of nonparametric inference (62G20) Nonparametric robustness (62G35) Multivariate analysis (62H99) Asymptotic distribution theory in statistics (62E20)

Cites Work

• Weak convergence and empirical processes. With applications to statistics
• Approximation Theorems of Mathematical Statistics
• Title not available (Why is that?)
• Relations between Weak and Uniform Convergence of Measures with Applications
• Robust Statistics
• Breakdown properties of location estimates based on halfspace depth and projected outlyingness
• Finite sample breakdown points of projection based multivariate location and scatter statistics
• Title not available (Why is that?)
• Title not available (Why is that?)
• Title not available (Why is that?)
• Convergence of stochastic processes
• Breakdown points of affine equivariant estimators of multivariate location and covariance matrices
• Multivariate analysis by data depth: Descriptive statistics, graphics and inference. (With discussions and rejoinder)
• General notions of statistical depth function.
• Title not available (Why is that?)
• On a notion of data depth based on random simplices
• Projection-based depth functions and associated medians
• Title not available (Why is that?)
• Limit theorems for the simplicial depth
• Do robust estimators work with real data?
• The asymptotic distribution of the trimmed mean
• On Some Robust Estimates of Location
• Descriptive statistics for nonparametric models. II: Location
• The 1972 Wald Lecture Robust Statistics: A Review
• Asymptotics for multivariate trimming
• On the Stahel-Donoho estimator and depth-weighted means of multivariate data.
• Influence function and maximum bias of projection depth based estimators.
• Asymptotic distributions of the maximal depth estimators for regression and multivariate location
• Asymptotics for the Tukey depth process, with an application to a multivariate trimmed mean
• Robustness of weighted \(L^p\)-depth and \(L^p\)-median
• A note on Tyler's modification of the MAD for the Stahel-Donoho estimator
• The metrically trimmed mean as a robust estimator of location

Cited In (25)

• On depth measures and dual statistics. A methodology for dealing with general data
• Local mutual information for dissimilarity-based image segmentation
• A note on weak convergence of general halfspace depth trimmed means
• On the Tukey depth of an atomic measure
• K-sign depth: from asymptotics to efficient implementation
• Asymptotics of generalized depth-based spread processes and applications
• Trimmed and Winsorized transformed means based on a scaled deviation
• Data depths satisfying the projection property
• Data depth trimming counterpart of the classical \(t\) (or \(T^2\)) procedure
• Trimmed and Winsorized means based on a scaled deviation
• Choosing among notions of multivariate depth statistics
• Fast Robust Location and Scatter Estimation: A Depth-based Method
• High dimensional data analysis using multivariate generalized spatial quantiles
• Least sum of squares of trimmed residuals regression
• Non-asymptotic analysis and inference for an outlyingness induced winsorized mean
• Trimmed and winsorized standard deviations based on a scaled deviation
• Multivariate trimmed means based on the Tukey depth
• Depth-weighted means of noisy data: an application to estimating the average effect in heterogeneous panels
• Bootstrap estimation of the proportion of outliers in robust regression
• A new approach for the computation of halfspace depth in high dimensions
• Computing projection depth and its associated estimators
• The Tukey and the random Tukey depths characterize discrete distributions
• On data depth in infinite dimensional spaces
• Robustness of the deepest projection regression functional
• The random Tukey depth

This page was built for publication: Multidimensional trimming based on projection depth

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q869968)