The Bernstein-Orlicz norm and deviation inequalities

DOI10.1007/S00440-012-0455-YMaRDI QIDQ377523zbMATH OpenOpenAlexFDO

Authors Johannes Lederer, Sara Van De Geer

Publication date 6 November 2013

Published in Zeitschrift für Wahrscheinlichkeitstheorie und Verwandte Gebiete (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/1111.2450

zbMATH Keywords

empirical processBernstein's inequalityOrlicz normchaining along a treedeviation inequality

Mathematics Subject Classification ID

Large deviations (60F10) Inequalities; stochastic orderings (60E15)

Abstract: We introduce two new concepts designed for the study of empirical processes. First, we introduce a new Orlicz norm which we call the Bernstein-Orlicz norm. This new norm interpolates sub-Gaussian and sub-exponential tail behavior. In particular, we show how this norm can be used to simplify the derivation of deviation inequalities for suprema of collections of random variables. Secondly, we introduce chaining and generic chaining along a tree. These simplify the well-known concepts of chaining and generic chaining. The supremum of the empirical process is then studied as a special case. We show that chaining along a tree can be done using entropy with bracketing. Finally, we establish a deviation inequality for the empirical process for the unbounded case.

Recommendations

• The Bennett-Orlicz norm
• A Bennett concentration inequality and its application to suprema of empirical processes
• A Bernstein-type inequality for suprema of random processes with applications to model selection in non-Gaussian regression
• A tail inequality for suprema of unbounded empirical processes with applications to Markov chains
• Tail bounds via generic chaining

Cites work

• scientific article; zbMATH DE number 3155055 (Why is no real title available?)
• scientific article; zbMATH DE number 5654889 (Why is no real title available?)
• A Bennett concentration inequality and its application to suprema of empirical processes
• A central limit theorem under metric entropy with \(L_ 2\) bracketing
• A tail inequality for suprema of unbounded empirical processes with applications to Markov chains
• About the constants in Talagrand's concentration inequalities for empirical processes.
• New concentration inequalities for suprema of empirical processes
• New concentration inequalities in product spaces
• On the Convergence of Empiric Distribution Functions
• Probability Inequalities for the Sum of Independent Random Variables
• Statistics for high-dimensional data. Methods, theory and applications.
• Supremum concentration inequality and modulus of continuity for sub-\(n\)th chaos processes
• The Generic Chaining
• The sizes of compact subsets of Hilbert space and continuity of Gaussian processes
• Weak convergence and empirical processes. With applications to statistics

Cited in

(22)

• Sharp large deviation results for sums of independent random variables
• Modified log-Sobolev inequalities and two-level concentration
• A Bernstein-type inequality for suprema of random processes with applications to model selection in non-Gaussian regression
• Some notes on concentration for \(\alpha\)-subexponential random variables
• Sharp oracle inequalities for low-complexity priors
• Cross-validation with confidence
• Prediction error bounds for linear regression with the TREX
• scientific article; zbMATH DE number 5885933 (Why is no real title available?)
• Exponential concentration inequalities for additive functionals of Markov chains
• Simultaneous phase retrieval and blind deconvolution via convex programming
• Statistical guarantees for regularized neural networks
• Concentration inequalities for separately convex functions
• Exponential concentration for geometric-median-of-means in non-positive curvature spaces
• Stochastic online convex optimization. Application to probabilistic time series forecasting
• The Bennett-Orlicz norm
• New concentration inequalities for suprema of empirical processes
• Functional linear regression with points of impact
• Concentration inequalities for suprema of unbounded empirical processes
• A robust, adaptive M-estimator for pointwise estimation in heteroscedastic regression
• scientific article; zbMATH DE number 7370543 (Why is no real title available?)
• Consistency of invariance-based randomization tests
• scientific article; zbMATH DE number 816272 (Why is no real title available?)

This page was built for publication: The Bernstein-Orlicz norm and deviation inequalities

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q377523)