Localized Gaussian width of \(M\)-convex hulls with applications to Lasso and convex aggregation

DOI10.3150/18-BEJ1078zbMath1428.62328arXiv1705.10696MaRDI QIDQ2325349

Pierre C. Bellec

Publication date: 25 September 2019

Published in: Bernoulli (Search for Journal in Brave)

Full work available at URL: https://arxiv.org/abs/1705.10696

zbMATH Keywords

convex hullLassoconvex aggregationGaussian mean widthanisotropic designlocalized Gaussian width

Mathematics Subject Classification ID

Nonparametric regression and quantile regression (62G08) Density estimation (62G07) Ridge regression; shrinkage estimators (Lasso) (62J07) Linear regression; mixed models (62J05)

Related Items (2)

ERM and RERM are optimal estimators for regression problems when malicious outliers corrupt the labels ⋮ On the robustness of minimum norm interpolators and regularized empirical risk minimizers

Cites Work

• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• A new perspective on least squares under convex constraint
• Optimal exponential bounds for aggregation of density estimators
• On the prediction performance of the Lasso
• Statistics for high-dimensional data. Methods, theory and applications.
• Exponential screening and optimal rates of sparse estimation
• Oracle inequalities in empirical risk minimization and sparse recovery problems. École d'Été de Probabilités de Saint-Flour XXXVIII-2008.
• Oracle inequalities and optimal inference under group sparsity
• Nuclear-norm penalization and optimal rates for noisy low-rank matrix completion
• Deviation optimal learning using greedy \(Q\)-aggregation
• Generalized mirror averaging and \(D\)-convex aggregation
• Linear and convex aggregation of density estimators
• Near-ideal model selection by \(\ell _{1}\) minimization
• Aggregation via empirical risk minimization
• The sparsity and bias of the LASSO selection in high-dimensional linear regression
• Learning by mirror averaging
• Fast learning rates for plug-in classifiers
• On the prediction loss of the Lasso in the partially labeled setting
• Bounds on the prediction error of penalized least squares estimators with convex penalty
• Optimal bounds for aggregation of affine estimators
• Sharp oracle inequalities for least squares estimators in shape restricted regression
• Persistene in high-dimensional linear predictor-selection and the virtue of overparametrization
• Mixing strategies for density estimation.
• Asymptotics for Lasso-type estimators.
• \(\ell _{1}\)-regularized linear regression: persistence and oracle inequalities
• Slope meets Lasso: improved oracle bounds and optimality
• Local Rademacher complexities and oracle inequalities in risk minimization. (2004 IMS Medallion Lecture). (With discussions and rejoinder)
• Simultaneous analysis of Lasso and Dantzig selector
• Empirical risk minimization is optimal for the convex aggregation problem
• Optimal learning with \textit{Q}-aggregation
• Gaussian averages of interpolated bodies and applications to approximate reconstruction
• Empirical minimization
• Local Rademacher complexities
• Learning without Concentration
• The Generalized Lasso With Non-Linear Observations
• Reconstruction From Anisotropic Random Measurements
• Scaled sparse linear regression
• Information Theory and Mixing Least-Squares Regressions
• High-dimensional estimation with geometric constraints: Table 1.
• Learning Theory and Kernel Machines
• Aggregation by Exponential Weighting and Sharp Oracle Inequalities
• Introduction to nonparametric estimation
• Sparse estimation by exponential weighting

This page was built for publication: Localized Gaussian width of \(M\)-convex hulls with applications to Lasso and convex aggregation