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.

Bayesian inference using least median of squares and least trimmed squares in models with independent or correlated errors and outliers

Robust Bayesian Inference for Berkson and Classical Measurement Error Models

Bayesian Restricted Likelihood Methods: Conditioning on Insufficient Statistics in Bayesian Regression

Bayesian estimation for longitudinal data in a joint model with HPCs

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

Empirical-Likelihood-Based Inference for Partially Linear Models

Maximum Likelihood Inference in Robust Linear Mixed-Effects Models Using Multivariate T Distributions

Generalized Median of Means Principle for Bayesian Inference

Bayesian analysis of linear regression models with autoregressive symmetrical errors and incomplete data

posteriordb: Testing, Benchmarking and Developing Bayesian Inference Algorithms

Robust Inference and Model Criticism Using Bagged Posteriors

Some Case Studies Using Bayesian Statistical Models

Robust Liu Estimator Used to Combat Some Challenges in Partially Linear Regression Model by Improving LTS Algorithm Using Semidefinite Programming

Robust Bayesian Model Averaging for Linear Regression Models With Heavy-Tailed Errors

Fast Bayesian inference in a class of sparse linear mixed effects models

Robust Bayesian nonparametric variable selection for linear regression

Reproducible Parameter Inference Using Bagged Posteriors

Fast Laplace Approximation for Sparse Bayesian Spike and Slab Models

Generalised Bayes Linear Inference

Bayesian Interpolation

A Bayesian approach to functional regression: theory and computation