On masking and swamping robustness of leading nonparametric outlier identifiers for multivariate data

Abstract For any outlier detection procedure, a key concern is robustness with respect to possible misclassification errors, masking (Type I) and swamping (Type II). Although parametric model-based simulation results are informative, one also desires nonparametric masking and robustness measures that are more broadly applicable. To this effect, notions of finite-sample masking and swamping breakdown points formulated abstractly for outlyingness functions in arbitrary data settings (Serfling and Wang, 2014) are introduced in the present paper into the multivariate data setting. Formulas for the measures are derived for three important affine invariant nonparametric multivariate outlyingness functions: Mahalanobis distance, Mahalanobis spatial, and projection. Using the formulas, favorable masking and swamping breakdown points, balanced equally, are seen for the Mahalanobis distance outlyingness using minimum covariance determinant (MCD) location and scatter estimators, and likewise for the projection outlyingness with median and MAD for univariate location and scale. Also, Mahalanobis spatial outlyingness with MCD standardization is competitive when swamping robustness is given higher priority than masking robustness. A small simulation study with bivariate contaminated standard normal and contaminated exponential models yields results consistent with the theoretical formulas. Some practical recommendations are discussed.

• Publication year

  2018

• Venue

  Journal of Multivariate Analysis

• Publication date

  2018-07-01

• Fields of study

  Mathematics, Computer Science

• Identifiers
  DOI 10.1016/j.jmva.2018.02.003
• External record

  Open on Semantic Scholar

• Source metadata

  Semantic Scholar

• No claims are published for this paper.

• No concepts are published for this paper.

• On masking and swamping robustness of leading nonparametric outlier identifiers for univariate data

  2015cited by this paper
• CompPD: A MATLAB Package for Computing Projection Depth

  2015cited by this paper
• General foundations for studying masking and swamping robustness of outlier identifiers

  2014cited by this paper
• Computationally easy outlier detection via projection pursuit with finitely many directions

  2013cited by this paper
• Computing projection depth and its associated estimators

  2012cited by this paper
• Equivariance and invariance properties of multivariate quantile and related functions, and the role of standardisation

  2010cited by this paper
• The forward search: theory and data analysis

  2010cited by this paper
• Multivariate Nonparametric Methods with R: An approach based on spatial signs and ranks

  2010cited by this paper
• Nonparametric depth-based multivariate outlier identifiers, and masking robustness properties

  2010cited by this paper
• Exponential probability inequality and convergence results for the median absolute deviation and its modifications

  2009cited by this paper
• Robust location and scatter estimators in multivariate analysis

  2006cited by this paper
• Depth weighted scatter estimators

  2005cited by this paper
• Projection-based depth functions and associated medians

  2003cited by this paper
• Robust Diagnostic Regression Analysis

  2002cited by this paper
• The multivariate L1-median and associated data depth.

  2000cited by this paper
• General notions of statistical depth function

  2000cited by this paper
• A Fast Algorithm for the Minimum Covariance Determinant Estimator

  1999cited by this paper
• The Masking Breakdown Point of Multivariate Outlier Identification Rules

  1999cited by this paper
• A note on Tyler's modification of the MAD for the Stahel-Donoho estimator

  1997cited by this paper
• Outliers in Statistical Data (3rd edition)

  1995influential reference
• Finite Sample Breakdown Points of Projection Based Multivariate Location and Scatter Statistics

  1994cited by this paper
• A Quality Index Based on Data Depth and Multivariate Rank Tests

  1993cited by this paper
• The identification of multiple outliers

  1993cited by this paper
• Breakdown Properties of Location Estimates Based on Halfspace Depth and Projected Outlyingness

  1992cited by this paper
• Asymptotic behaviour of S-estimates of multivariate location parameters and dispersion matrices

  1987cited by this paper
• Multivariate estimation with high breakdown point

  1985cited by this paper
• Outliers in Statistical Data

  1979cited by this paper
• Data Analysis and Regression

  1978cited by this paper
• On the generalized distance in statistics

  1936cited by this paper

• On the robustness of spatial quantiles

  2026cites this paper
• A data-driven approach to study temporal characteristics of COVID-19 infection and death Time Series for twelve countries across six continents

  2025cites this paper
• Finite Sample Breakdown Point of the Minimum Volume Ellipsoid Estimator in Extended General Position of Sample

  2023cites this paper
• Improved Spatial Outlier Detection Method within a River Network

  2022cites this paper
• Detecting Outliers in Data Streams Based on Minimum Rare Pattern Mining and Pattern Matching

  2022influential citation
• An Improved Weighted Median Algorithm for Spatial Outliers Detection

  2022cites this paper
• Competitiveness Scale as a Basis for Brazilian Small and Medium-Sized Enterprises

  2020cites this paper
• D ec 2 01 8 On general notions of depth for regression

  2018cites this paper
• On General Notions of Depth for Regression

  2018cites this paper
• Perspectives on Depth Functions on General Data Spaces , with Consideration of the Tukey , Projection , Spatial , “ Density ” , “ Local ” , and “ Contour ” Depths

  year unknowncites this paper
• Depth Functions on General Data Spaces, II. Formulation and Maximality, with Consideration of the Tukey, Projection, Spatial, and “Contour” Depths

  year unknowncites this paper