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.

Large sample properties of maximum likelihood estimator using moving extremes ranked set sampling

Double Moving Extremes Ranked Set Sampling Design

A Comparison of MLE for Some Index Distributions Based on Censored Samples

Maximum likelihood estimation of the parameters of the inverse Gaussian distribution using maximum rank set sampling with unequal samples

Weighted exponential parameters estimation using maximum ranked set sampling with unequal samples

Tilting maximum Lq-Likelihood estimation for extreme values drawing on block maxima

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

On the MLE of the Waring Distribution

Estimating Extreme Value Index by Subsampling for Massive Datasets with Heavy-Tailed Distributions

On an Asymptotic Distribution for the MLE

Maximum likelihood estimators based on the block maxima method

Random pairing MLE for estimation of item parameters in Rasch model

Maximum Likelihood Estimation of Flexible Survival Densities with Importance Sampling

Finite-sample performance of the maximum likelihood estimator in logistic regression

Near Universal Consistency of the Maximum Pseudolikelihood Estimator for Discrete Models

Sampling in Latent Space for a Mulitiobjective Estimation of Distribution Algorithm

Robust estimators for the log-logistic model based on ranked set sampling

Maximum Likelihood Estimations and EM Algorithms with Length-biased Data

Expectation-Maximization-Maximization: A Feasible MLE Algorithm for the Three-Parameter Logistic Model Based on a Mixture Modeling Reformulation

Semiparametric maximum likelihood estimation with two-phase stratified case-control sampling

Maximum-Likelihood-Estimation of Parameters under A Spatial Sampling Scheme