Optimal rank-based tests for homogeneity of scatter

DOI10.1214/07-AOS508zbMATH Open1360.62288arXiv0806.2963OpenAlexW3098345134MaRDI QIDQ930654FDOQ930654

Authors: Marc Hallin, Davy Paindaveine 

Publication date: 1 July 2008

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

Abstract: We propose a class of locally and asymptotically optimal tests, based on multivariate ranks and signs for the homogeneity of scatter matrices in $m$ elliptical populations. Contrary to the existing parametric procedures, these tests remain valid without any moment assumptions, and thus are perfectly robust against heavy-tailed distributions (validity robustness). Nevertheless, they reach semiparametric efficiency bounds at correctly specified elliptical densities and maintain high powers under all (efficiency robustness). In particular, their normal-score version outperforms traditional Gaussian likelihood ratio tests and their pseudo-Gaussian robustifications under a very broad range of non-Gaussian densities including, for instance, all multivariate Student and power-exponential distributions.

Full work available at URL: https://arxiv.org/abs/0806.2963

Recommendations

• Optimal tests for homogeneity of covariance, scale, and shape
• Semiparametrically efficient rank-based inference for shape. I: optimal rank-based tests for sphericity
• Optimal rank-based testing for principal components
• Optimal tests for multivariate location based on interdirections and pseudo-Mahalanobis ranks.
• Robustness and efficiency properties of scatter matrices

zbMATH Keywords

semiparametric efficiencyadaptivitylocal asymptotic normalityscatter matrixshape matrixelliptic densities

Mathematics Subject Classification ID

Nonparametric hypothesis testing (62G10) Nonparametric robustness (62G35) Hypothesis testing in multivariate analysis (62H15)

Cites Work

• Distribution-Free Two-Sample Tests for Scale
• Title not available (Why is that?)
• Properties of sufficiency and statistical tests
• A distribution-free M-estimator of multivariate scatter
• On the uniqueness of \(S\)-functionals and \(M\)-functionals under nonelliptical distributions.
• Semiparametrically efficient rank-based inference for shape. I: optimal rank-based tests for sphericity
• Semiparametrically efficient rank-based inference for shape. II: Optimal \(R\)-estimation of shape
• Principal Components Analysis Based on Multivariate MM Estimators With Fast and Robust Bootstrap
• Title not available (Why is that?)
• Title not available (Why is that?)
• Asymptotic methods in statistical decision theory
• On adaptive estimation
• On some tests of the covariance matrix under general conditions
• Asymptotic Normality of Simple Linear Rank Statistics Under Alternatives
• A practical affine equivariant multivariate median
• Some tests for the equality of covariance matrices
• A canonical definition of shape
• Title not available (Why is that?)
• Title not available (Why is that?)
• Robust Bounded-Influence Tests in General Parametric Models
• Title not available (Why is that?)
• Asymptotic Linearity of a Rank Statistic in Regression Parameter
• On some test criteria for covariance matrix
• A Simpler, Affine-Invariant, Multivariate, Distribution-Free Sign Test
• On Tyler's \(M\)-functional of scatter in high dimension
• Optimal tests for multivariate location based on interdirections and pseudo-Mahalanobis ranks.
• Title not available (Why is that?)
• A Bootstrap Comparison of Genetic Covariance Matrices
• On adaptive estimation in stationary ARMA processes
• Unbiasedness of the likelihood ratio tests for equality of several covariance matrices and equality of several multivariate normal populations
• Rank-based optimal tests of the adequacy of an elliptic VARMA model
• The effects of nonnormality on asymptotic distributions of some likelihood ratio criteria for testing covariance structures under normal assumption
• Influence functions and efficiencies of the canonical correlation and vector estimates based on scatter and shape matrices
• On the Breakdown Properties of Some Multivariate M-Functionals*
• NON-NORMALITY AND TESTS ON VARIANCES
• Semi-parametric efficiency, distribution-freeness and invariance
• Estimates of Regression Coefficients Based on the Sign Covariance Matrix
• Robustness and efficiency properties of scatter matrices
• Parametric and semiparametric inference for shape: the role of the scale functional
• A Chernoff-Savage result for shape: On the non-admissibility of pseudo-Gaussian methods
• Multivariate signed-rank tests in vector autoregressive order identification
• Bootstrap Critical Values for Testing Homogeneity of Covariance Matrices

Cited In (15)

• Optimal tests for elliptical symmetry: specified and unspecified location
• Asymptotic distributions of robust shape matrices and scales
• Rank-based testing for semiparametric VAR models: a measure transportation approach
• Semiparametrically efficient rank-based inference for shape. I: optimal rank-based tests for sphericity
• Optimal rank-based testing for principal components
• Distribution and quantile functions, ranks and signs in dimension \(d\): a measure transportation approach
• Rank Test of Location Optimal for Hyperbolic Secant Distribution
• A conversation with Marc Hallin
• Optimal tests for homogeneity of covariance, scale, and shape
• On Hodges and Lehmann’s “6/π Result”
• Semiparametric error-correction models for cointegration with trends: pseudo-Gaussian and optimal rank-based tests of the cointegration rank
• Robust tests for scatter separability beyond Gaussianity
• An exact rank test for scale under normality using Helmert's transformation
• A canonical definition of shape
• Optimal rank-based tests for common principal components

This page was built for publication: Optimal rank-based tests for homogeneity of scatter

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q930654)