Abstract:Heavy-tailed errors impair the accuracy of the least squares estimate, which can be spoiled by a single grossly outlying observation. As argued in the seminal work of Peter Huber in 1973 [Ann. Statist.1 (1973) 799-821], robust alternatives to the method of least squares are sorely needed. To achieve robustness against heavy-tailed sampling distributions, we revisit the Huber estimator from a new perspective by letting the tuning parameter involved diverge with the sample size. In this paper, we develop nonasymptotic concentration results for such an adaptive Huber estimator, namely, the Huber estimator with the tuning parameter adapted to sample size, dimension, and the variance of the noise. Specifically, we obtain a sub-Gaussian-type deviation inequality and a nonasymptotic Bahadur representation when noise variables only have finite second moments. The nonasymptotic results further yield two conventional normal approximation results that are of independent interest, the Berry-Esseen inequality and Cramér-type moderate deviation. As an important application to large-scale simultaneous inference, we apply these robust normal approximation results to analyze a dependence-adjusted multiple testing procedure for moderately heavy-tailed data. It is shown that the robust dependence-adjusted procedure asymptotically controls the overall false discovery proportion at the nominal level under mild moment conditions. Thorough numerical results on both simulated and real datasets are also provided to back up our theory.

Asymptotics of nonparametric L-1 regression models with dependent data

Asymptotic normality for L1-norm kernel estimator of conditional median under association dependence

Asymptotic Theory of Least Distances Estimate in Multivariate Linear Models

ASYMPTOTIC NORMALITY OF KERNEL ESTIMATES OF A DENSITY FUNCTION UNDER ASSOCIATION DEPENDENCE

Asymptotics for A Class of Dependent Random Variables

Flexible Bayesian quantile regression for nonlinear mixed effects models based on the generalized asymmetric Laplace distribution

Asymptotics of Linear Regression with Linearly Dependent Data

Local Linear Quantile Regression for Time Series Under Near Epoch Dependence

Asymptotic Properties of Nonparametric M-Estimation for Mixing Functional Data

A NEW PERSPECTIVE ON ROBUST M-ESTIMATION: FINITE SAMPLE THEORY AND APPLICATIONS TO DEPENDENCE-ADJUSTED MULTIPLE TESTING

Asymptotic Distributions of Nonparametric Regression Estimators for Longitudinal or Functional Data

Asymptotic normality of locally modelled regression estimator for functional data

On the Asymptotics of Marginal Regression Splines with Longitudinal Data

Asymptotic Normality in Partial Linear Models Based on Dependent Errors

Asymptotics of Estimates in Constrained Nonlinear Regression with Long-Range Dependent Innovations

Efficient L-1 Estimation and Related Inferences in Linear Regression with Unknown Form of Heteroscedasticity

The Asymptotic Normality of Internal Estimator for Nonparametric Regression.

On Degrees of Freedom of Projection Estimators With Applications to Multivariate Nonparametric Regression

Analysis of Least Absolute Deviation

Minimax rates of convergence for nonparametric location-scale models

Central limit theorems for high dimensional dependent data