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.

New Concentration Metrics for Performance Evaluation of Estimation Algorithms.

Simplified Desirability Level Metrics For Estimation Performance Evaluation

Optimized estimation for population mean using conventional and non-conventional measures under the joint influence of measurement error and non-response

New Robust Metrics of Central Tendency for Estimation Performance Evaluation

Improved Estimation of Population Mean Through Known Conventional and Non-Conventional Measures of Auxiliary Variable

Dynamic Error Spectrum for Estimation Performance Evaluation: a Case Study on Interacting Multiple Model Algorithm

On the Concentration of the Minimizers of Empirical Risks

Improved Estimation of Concentration Under $\ell_p$-Norm Distance Metrics Using Half Spaces

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

An Estimate of a Change Point in Variance of Measurement Errors and Its Convergence Rate

Dynamic error spectrum for IMM performance evaluation

Measures for Ranking Estimation Performance Based on Single or Multiple Performance Metrics

Enhanced Error Spectrum for Estimation Performance Evaluation

Convergence Analysis on Data-Driven Fortet-Mourier Metrics with Applications in Stochastic Optimization

A Concentration-Based Approach for Optimizing the Estimation Performance in Stochastic Sensor Selection

Performance Limits with Additive Error Metrics in Noisy Multi-Measurement Vector Problem

Concentration Inequalities for Statistical Inference

A Unified Approach to Focused Information Criterion and Plug-In Averaging Method

Achieving High Convergence Rates by Quasi-Monte Carlo and Importance Sampling for Unbounded Integrands

On Estimation and Optimization of Mean Values of Bounded Variables

Concentration inequalities for locally small increments of compound empirical processes with applications to solutions of compound and risk averse stochastical programming