An improved global risk bound in concave regression

DOI10.1214/16-EJS1151zbMATH Open1349.62126MaRDI QIDQ309525FDOQ309525

Authors Sabyasachi Chatterjee

Publication date 7 September 2016

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

Full work available at URL https://arxiv.org/abs/1512.04658

zbMATH Keywords

concave regressionglobal risk bound

Mathematics Subject Classification ID

Nonparametric regression and quantile regression (62G08) Asymptotic properties of nonparametric inference (62G20)

Abstract: A new risk bound is presented for the problem of convex/concave function estimation, using the least squares estimator. The best known risk bound, as had appeared in citet{GSvex}, scaled like $log(en)n^{-4/5}$ under the mean squared error loss, up to a constant factor. The authors in cite{GSvex} had conjectured that the logarithmic term may be an artifact of their proof. We show that indeed the logarithmic term is unnecessary and prove a risk bound which scales like $n^{-4/5}$ up to constant factors. Our proof technique has one extra peeling step than in a usual chaining type argument. Our risk bound holds in expectation as well as with high probability and also extends to the case of model misspecification, where the true function may not be concave.

Recommendations

• Global risk bounds and adaptation in univariate convex regression
• scientific article; zbMATH DE number 4056794
• A new concentration result for regularized risk minimizers
• On concentration for (regularized) empirical risk minimization
• On the risk of convex-constrained least squares estimators under misspecification
• On improved shrinkage estimators for concave loss
• Concave regression: value-constrained estimation and likelihood ratio-based inference
• A risk bound in Sobolev class regression
• Minimum Convex Risk Equivariant Finite Population Prediction for Linear Functions in Regression Models

Cites work

• scientific article; zbMATH DE number 193111 (Why is no real title available?)
• scientific article; zbMATH DE number 1420699 (Why is no real title available?)
• A canonical process for estimation of convex functions: the ``invelope of integrated Brownian motion \(+t^ 4\).
• A new perspective on least squares under convex constraint
• Adaptive estimation of a quadratic functional by model selection.
• An Empirical Distribution Function for Sampling with Incomplete Information
• Consistency in concave regression
• Consistency of concave regression with an application to current-status data
• Global risk bounds and adaptation in univariate convex regression
• Kolmogorov entropy for classes of convex functions
• Nonparametric regression under qualitative smoothness assumptions
• On risk bounds in isotonic and other shape restricted regression problems
• Point Estimates of Ordinates of Concave Functions
• Risk bounds in isotonic regression

Cited in

(9)

• Nonparametric shape-restricted regression
• Sharp oracle inequalities for least squares estimators in shape restricted regression
• Global risk bounds and adaptation in univariate convex regression
• On risk bounds in isotonic and other shape restricted regression problems
• Stratified incomplete local simplex tests for curvature of nonparametric multiple regression
• scientific article; zbMATH DE number 6617259 (Why is no real title available?)
• On the risk of convex-constrained least squares estimators under misspecification
• The limiting behavior of isotonic and convex regression estimators when the model is misspecified
• Convexification for data fitting

This page was built for publication: An improved global risk bound in concave regression

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q309525)