Nonparametric Imputation by Data Depth

DOI10.1080/01621459.2018.1543123MaRDI QIDQ3304851zbMATH OpenOpenAlexWikidataFDO

Authors Pavlo Mozharovskyi, Julie Josse, François Husson

Publication date 3 August 2020

Published in Journal of the American Statistical Association (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/1701.03513

zbMATH Keywords

convex optimizationelliptical symmetrylocal depthnonparametric imputationoutliersTukey depthzonoid depth

Mathematics Subject Classification ID

Software, source code, etc. for problems pertaining to statistics (62-04) Missing data (62D10) Linear programming (90C05) Convex programming (90C25)

Abstract: We present single imputation method for missing values which borrows the idea of data depth---a measure of centrality defined for an arbitrary point of a space with respect to a probability distribution or data cloud. This consists in iterative maximization of the depth of each observation with missing values, and can be employed with any properly defined statistical depth function. For each single iteration, imputation reverts to optimization of quadratic, linear, or quasiconcave functions that are solved analytically by linear programming or the Nelder-Mead method. As it accounts for the underlying data topology, the procedure is distribution free, allows imputation close to the data geometry, can make prediction in situations where local imputation (k-nearest neighbors, random forest) cannot, and has attractive robustness and asymptotic properties under elliptical symmetry. It is shown that a special case---when using the Mahalanobis depth---has direct connection to well-known methods for the multivariate normal model, such as iterated regression and regularized PCA. The methodology is extended to multiple imputation for data stemming from an elliptically symmetric distribution. Simulation and real data studies show good results compared with existing popular alternatives. The method has been implemented as an R-package. Supplementary materials for the article are available online.

Recommendations

• Nonparametric conditional mean imputation
• Local multiple imputation
• A numerical study of multiple imputation methods using nonparametric multivariate outlier identifiers and depth-based performance criteria with clinical laboratory data

Cites work

• A general solution for robust linear programs with distortion risk constraints
• A Quality Index Based on Data Depth and Multivariate Rank Tests
• Breakdown properties of location estimates based on halfspace depth and projected outlyingness
• Bridging centrality and extremity: refining empirical data depth using extreme value statistics
• Clustering and classification based on the L\(_{1}\) data depth
• Data depths satisfying the projection property
• Exact computation of the halfspace depth
• Fast nonparametric classification based on data depth
• Flexible imputation of missing data.
• From Depth to Local Depth: A Focus on Centrality
• General notions of statistical depth function.
• scientific article; zbMATH DE number 1818892 (Why is no real title available?)
• scientific article; zbMATH DE number 45785 (Why is no real title available?)
• scientific article; zbMATH DE number 3541764 (Why is no real title available?)
• scientific article; zbMATH DE number 3567782 (Why is no real title available?)
• scientific article; zbMATH DE number 1294360 (Why is no real title available?)
• scientific article; zbMATH DE number 1834445 (Why is no real title available?)
• scientific article; zbMATH DE number 6458321 (Why is no real title available?)
• scientific article; zbMATH DE number 3023295 (Why is no real title available?)
• Matrix completion and low-rank SVD via fast alternating least squares
• Missing Data, Imputation, and the Bootstrap
• Monotonicity properties of spatial depth
• Multivariate analysis by data depth: Descriptive statistics, graphics and inference. (With discussions and rejoinder)
• Multivariate dispersion, central regions and depth. The lift zonoid approach
• Multivariate risks and depth-trimmed regions
• Statistical Applications of the Multivariate Skew Normal Distribution
• Structural properties and convergence results for contours of sample statistical depth functions.
• The multivariate L ₁ -median and associated data depth
• Zonoid trimming for multivariate distributions

Cited in

(8)

• Exploring copulas for the imputation of complex dependent data
• Imputation in nonparametric quantile regression with complex data
• Imputing continuous data under some non‐Gaussian distributions
• Depth-based reconstruction method for incomplete functional data
• Inference in coarsened time series via generalized method of moments
• A novel test of missing completely at random: U -statistics-based approach
• Halfspace depth
• On a Notion of Graph Centrality Based on L ₁ Data Depth

Describes a project that uses

Uses Software

• R
• missForest
• impute

This page was built for publication: Nonparametric Imputation by Data Depth

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q3304851)