Influence functions for penalized M-estimators

DOI10.3150/16-BEJ841MaRDI QIDQ2405153zbMATH OpenOpenAlexFDO

Authors Marco Avella-Medina

Publication date 21 September 2017

Published in Bernoulli (Search for Journal in Brave)

Full work available at URL https://projecteuclid.org/euclid.bj/1495505089

zbMATH Keywords

distribution theoryimplicit function theoremLassoregularizationrobust statistics

Mathematics Subject Classification ID

Robustness and adaptive procedures (parametric inference) (62F35) Ridge regression; shrinkage estimators (Lasso) (62J07)

Review text

The classical definition of influence function for a regular M-estimator is useful in the assessment of relative influence of individual observations towards the value of an estimate and its asymptotic properties of an estimate. This definition is, however, difficult to apply to penalized M-estimators which are obtained by minimizing a regularized risk since the penalty function in the definition of the risk function may not be differentiable. In this paper, a new definition of influence function is proposed as the limit of the influence functions for a sequence of penalized M-estimators defined by a sequence of smooth penalty functions which approach the potentially non-smooth penalty function. It is shown that this limit is uniquely defined which is unrelated to the chosen limiting sequence. This definition has been used to derive the influence functions of penalized M-estimators defined by Lasso and group Lasso type penalties and extended two-stage penalized M-estimators such as those defined by adaptive Lasso type penalties.

Recommendations

• The influence function of penalized regression estimators
• Robust and sparse estimators for linear regression models
• Generalized influence functions and robustness analysis
• Penalised robust estimators for sparse and high-dimensional linear models
• Estimation of influence functions

Cited in

(14)

• Wavelet-based robust estimation and variable selection in nonparametric additive models
• Scale calibration for high-dimensional robust regression
• Asymptotic linear expansion of regularized M-estimators
• The influence function of penalized regression estimators
• Privacy-preserving parametric inference: a case for robust statistics
• Estimators of Influence Function
• Generalized influence functions and robustness analysis
• The robust desparsified lasso and the focused information criterion for high-dimensional generalized linear models
• Gaining Outlier Resistance With Progressive Quantiles: Fast Algorithms and Theoretical Studies
• The influence function of semiparametric two-step estimators with estimated control variables
• Robust and consistent model evaluation criteria in high-dimensional regression
• High-dimensional outlier detection and variable selection via adaptive weighted mean regression
• Robust adaptive Lasso in high-dimensional logistic regression
• The main contributions of robust statistics to statistical science and a new challenge

This page was built for publication: Influence functions for penalized M-estimators

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2405153)