Nonparametrically consistent depth-based classifiers

DOI10.3150/13-BEJ561zbMATH Open1359.62258arXiv1204.2996MaRDI QIDQ2345115FDOQ2345115

Authors: Davy Paindaveine, Germain Van Bever 

Publication date: 19 May 2015

Published in: Bernoulli (Search for Journal in Brave)

Abstract: We introduce a class of depth-based classification procedures that are of a nearest-neighbor nature. Depth, after symmetrization, indeed provides the center-outward ordering that is necessary and sufficient to define nearest neighbors. Like all their depth-based competitors, the resulting classifiers are affine-invariant, hence in particular are insensitive to unit changes. Unlike the former, however, the latter achieve Bayes consistency under virtually any absolutely continuous distributions - a concept we call nonparametric consistency, to stress the difference with the stronger universal consistency of the standard $k$NN classifiers. We investigate the finite-sample performances of the proposed classifiers through simulations and show that they outperform affine-invariant nearest-neighbor classifiers obtained through an obvious standardization construction. We illustrate the practical value of our classifiers on two real data examples. Finally, we shortly discuss the possible uses of our depth-based neighbors in other inference problems.

Full work available at URL: https://arxiv.org/abs/1204.2996

Recommendations

• Affine-invariant depth-based classifiers on the basis of the \(k\)-nearest neighbours method
• Research of nonparametric data classifiers based on the depth neighbourhoods concept
• A depth-based modification of the \(k\)-nearest neighbour method.
• Depth-weighted Bayes classification
• On Maximum Depth and Related Classifiers

zbMATH Keywords

universal consistencyabsolutely continuous distributions\(k\)NN classifiersnonparametric consistency

Mathematics Subject Classification ID

Asymptotic properties of nonparametric inference (62G20) Classification and discrimination; cluster analysis (statistical aspects) (62H30)

Cites Work

• Fast nonparametric classification based on data depth
• A distribution-free M-estimator of multivariate scatter
• Breakdown properties of location estimates based on halfspace depth and projected outlyingness
• Consistent nonparametric regression. Discussion
• High breakdown estimation for multiple populations with applications to discriminant analysis
• Robust linear discriminant analysis using S-estimators
• Title not available (Why is that?)
• Zonoid trimming for multivariate distributions
• Multivariate analysis by data depth: Descriptive statistics, graphics and inference. (With discussions and rejoinder)
• On data depth and distribution-free discriminant analysis using separating surfaces
• General notions of statistical depth function.
• Clustering and classification based on the L\(_{1}\) data depth
• On robust classification using projection depth
• Title not available (Why is that?)
• DD-classifier: nonparametric classification procedure based on DD-plot
• On Maximum Depth and Related Classifiers
• On a notion of data depth based on random simplices
• Projection-based depth functions and associated medians
• Asymptotics for general multivariate kernel density derivative estimators
• A practical affine equivariant multivariate median
• Limit theorems for the simplicial depth
• Structural properties and convergence results for contours of sample statistical depth functions.
• The depth function of a population distribution.
• On Tyler's \(M\)-functional of scatter in high dimension
• Characterizing angular symmetry and regression symmetry.
• Optimal signed-rank tests based on hyperplanes
• Robust classification for skewed data
• Consistency and localizability
• An affine invariant \(k\)-nearest neighbor regression estimate
• Generalized Linear and Quadratic Discriminant Functions Using Robust Estimates
• Data‐driven choice of the smoothing parametrization for kernel density estimators
• An Extended Projection Data Depth and Its Applications to Discrimination

Cited In (23)

• Affine-invariant depth-based classifiers on the basis of the \(k\)-nearest neighbours method
• Illumination Depth
• Research of nonparametric data classifiers based on the depth neighbourhoods concept
• Depth-based classification method underlain by a remote concentration measure for processing asymmetric data
• A topologically valid definition of depth for functional data
• Halfspace depths for scatter, concentration and shape matrices
• Asymptotic estimate of depth-based classifiers within the location shift model
• On similarity of the sample projection depth contours and its application
• On some classifiers based on multivariate ranks
• Title not available (Why is that?)
• On general notions of depth for regression
• Classifying real-world data with the \(DD\alpha\)-procedure
• Affine-invariant classifier of extrapolation depth on the basis of a multilevel smoothing structure
• \(k\)NN estimation in functional partial linear modeling
• A depth-based modification of the \(k\)-nearest neighbour method.
• \(k\)-depth-nearest neighbour method and its performance on skew-normal distributons
• Depth-weighted Bayes classification
• Depth-based classification for distributions with nonconvex support
• Dilemmas of robust analysis of economic data streams
• A weighted localization of halfspace depth and its properties
• RR-classifier: a nonparametric classification procedure in multidimensional space based on relative ranks
• Depth for curve data and applications
• Classification with the pot-pot plot

This page was built for publication: Nonparametrically consistent depth-based classifiers

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2345115)