Robust dimension-free Gram operator estimates

DOI10.3150/17-BEJ981MaRDI QIDQ1750105zbMATH OpenOpenAlexFDO

Authors Ilaria Giulini

Publication date 18 May 2018

Published in Bernoulli (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/1511.06259, https://projecteuclid.org/euclid.bj/1524038772

zbMATH Keywords

dimension-free boundsGram operatorPAC-Bayesian learningrobust estimation

Mathematics Subject Classification ID

Asymptotic distribution theory in statistics (62E20) Nonparametric estimation (62G05) Nonparametric robustness (62G35) Estimation in multivariate analysis (62H12)

Abstract: In this paper we investigate the question of estimating the Gram operator by a robust estimator from an i.i.d. sample in a separable Hilbert space and we present uniform bounds that hold under weak moment assumptions. The approach consists in first obtaining non-asymptotic dimension-free bounds in finite-dimensional spaces using some PAC-Bayesian inequalities related to Gaussian perturbations of the parameter and then in generalizing the results in a separable Hilbert space. We show both from a theoretical point of view and with the help of some simulations that such a robust estimator improves the behavior of the classical empirical one in the case of heavy tail data distributions.

Recommendations

• Robust PCA and pairs of projections in a Hilbert space
• Robust estimation of dimension reduction space
• Concentration study of M-estimators using the influence function
• All-in-one robust estimator of the Gaussian mean
• Geometric median and robust estimation in Banach spaces

Cites work

• Asymptotics and concentration bounds for bilinear forms of spectral projectors of sample covariance
• Challenging the empirical mean and empirical variance: a deviation study
• Consistency of spectral clustering
• Geometric median and robust estimation in Banach spaces
• scientific article; zbMATH DE number 1966603 (Why is no real title available?)
• Learning Theory
• On learning with integral operators
• On the Eigenspectrum of the Gram Matrix and the Generalization Error of Kernel-PCA
• Random matrix approximation of spectra of integral operators
• Random vectors in the isotropic position
• User-friendly tail bounds for sums of random matrices

Cited in

(9)

• Finite sample properties of parametric MMD estimation: robustness to misspecification and dependence
• User-friendly Introduction to PAC-Bayes Bounds
• Dimension-free bounds for sums of independent matrices and simple tensors via the variational principle
• Mean estimation in high dimension
• Dimension-free bounds for sums of dependent matrices and operators with heavy-tailed distributions
• Covariance estimation: optimal dimension-free guarantees for adversarial corruption and heavy tails
• Uniform bounds for robust mean estimators
• Statistically optimal robust mean and covariance estimation for anisotropic gaussians
• Dimension-free uniform concentration bound for logistic regression

This page was built for publication: Robust dimension-free Gram operator estimates

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q1750105)