Adaptive spectral regularizations of high dimensional linear models

DOI10.1214/11-EJS649MaRDI QIDQ1952239zbMATH OpenFDO

Authors Yu. Golubev

Publication date 28 May 2013

Published in Electronic Journal of Statistics (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/1112.5890, https://projecteuclid.org/euclid.ejs/1322057437

zbMATH Keywords

empirical risk minimizationexcess riskoracle inequalityordered smoother

Mathematics Subject Classification ID

Estimation in multivariate analysis (62H12) Linear regression; mixed models (62J05)

Abstract: This paper focuses on recovering an unknown vector $$ from the noisy data $$, where $X$ is a known $nimesp$-matrix, $xi$ is a standard white Gaussian noise, and $sigma$ is an unknown noise level. In order to estimate $$, a spectral regularization method is used, and our goal is to choose its regularization parameter with the help of the data $Y$. In this paper, we deal solely with regularization methods based on the so-called ordered smoothers and provide some oracle inequalities in the case, where the noise level is unknown.

Recommendations

• On universal oracle inequalities related to high-dimensional linear models
• On oracle inequalities related to high dimensional linear models
• Spectral cut-off regularizations for ill-posed linear models
• Noise level estimation in high-dimensional linear models
• Tikhonov-Phillips regularizations in linear models with blurred design

Cites work

• A Lepskij-type stopping rule for regularized Newton methods
• A statistical perspective on ill-posed inverse problems (with discussion)
• Adaptive complexity regularization for linear inverse problems
• An Iteration Formula for Fredholm Integral Equations of the First Kind
• Convergence Rates of General Regularization Methods for Statistical Inverse Problems and Applications
• Fundamentals of Computerized Tomography
• scientific article; zbMATH DE number 1183934 (Why is no real title available?)
• scientific article; zbMATH DE number 3551792 (Why is no real title available?)
• scientific article; zbMATH DE number 3444596 (Why is no real title available?)
• scientific article; zbMATH DE number 936298 (Why is no real title available?)
• Local solutions to inverse problems in geodesy. The impact of the noise covariance structure upon the accuracy of estimation
• Minimal penalties for Gaussian model selection
• On universal oracle inequalities related to high-dimensional linear models
• Optimal filtering of square-integrable signals in Gaussian noise
• Oracle inequalities for inverse problems
• Ordered linear smoothers
• Risk hull method and regularization by projections of ill-posed inverse problems
• Statistical Inverse Estimation in Hilbert Scales
• The Lepskii principle revisited
• The principle of penalized empirical risk in severely ill-posed problems
• Weak convergence and empirical processes. With applications to statistics

Cited in

(8)

• Empirical risk minimization as parameter choice rule for general linear regularization methods
• Spectral cut-off regularizations for ill-posed linear models
• On risk concentration for convex combinations of linear estimators
• On oracle inequalities related to high dimensional linear models
• Optimal adaptation for early stopping in statistical inverse problems
• Adaptive k-class estimation in high-dimensional linear models
• On universal oracle inequalities related to high-dimensional linear models
• Tikhonov-Phillips regularizations in linear models with blurred design

This page was built for publication: Adaptive spectral regularizations of high dimensional linear models

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q1952239)