Confidence intervals for multiple isotonic regression and other monotone models

DOI10.1214/20-AOS2025zbMATH Open1478.62104arXiv2001.07064OpenAlexW3202543529MaRDI QIDQ2054474FDOQ2054474

Authors: Yanyan Li 

Publication date: 3 December 2021

Published in: The Annals of Statistics (Search for Journal in Brave)

Abstract: We consider the problem of constructing pointwise confidence intervals in the multiple isotonic regression model. Recently, [HZ19] obtained a pointwise limit distribution theory for the so-called block max-min and min-max estimators [FLN17] in this model, but inference remains a difficult problem due to the nuisance parameter in the limit distribution that involves multiple unknown partial derivatives of the true regression function. In this paper, we show that this difficult nuisance parameter can be effectively eliminated by taking advantage of information beyond point estimates in the block max-min and min-max estimators. Formally, let $hatu(x_0)$ (resp. $hatv(x_0)$) be the maximizing lower-left (resp. minimizing upper-right) vertex in the block max-min (resp. min-max) estimator, and $hat{f}_n$ be the average of the block max-min and min-max estimators. If all (first-order) partial derivatives of $f_0$ are non-vanishing at $x_0$, then the following pivotal limit distribution theory holds: ${\displaystyle }$ sqrt{n_{hat{u},hat{v}}(x_0)}�ig(hat{f}_n(x_0)-f_0(x_0)�ig) ightsquigarrow sigmacdot mathbb{L}_{1_d}. ${\displaystyle }$ Here $n_{hatu,hatv}(x_0)$ is the number of design points in the block $[hatu(x_0),hatv(x_0)]$, $sigma$ is the standard deviation of the errors, and $mathbb{L}_{1_d}$ is a universal limit distribution free of nuisance parameters. This immediately yields confidence intervals for $f_0(x_0)$ with asymptotically exact confidence level and oracle length. Notably, the construction of the confidence intervals, even new in the univariate setting, requires no more efforts than performing an isotonic regression for once using the block max-min and min-max estimators, and can be easily adapted to other common monotone models. Extensive simulations are carried out to support our theory.

Full work available at URL: https://arxiv.org/abs/2001.07064

Recommendations

• Nonparametric confidence intervals for monotone functions
• Limit distribution theory for block estimators in multiple isotonic regression
• scientific article; zbMATH DE number 1267279
• Adaptive confidence sets in shape restricted regression
• Monotone nonparametric regression and confidence intervals

zbMATH Keywords

confidence intervalGaussian processshape constraintslimit distribution theorymultiple isotonic regression

Mathematics Subject Classification ID

Nonparametric estimation (62G05) Nonparametric regression and quantile regression (62G08) Asymptotic distribution theory in statistics (62E20) Nonparametric tolerance and confidence regions (62G15)

Cites Work

• Inference for the mode of a log-concave density
• Generalized additive and index models with shape constraints
• Weak convergence and empirical processes. With applications to statistics
• Limit distribution theory for maximum likelihood estimation of a log-concave density
• Approximation by log-concave distributions, with applications to regression
• Maximum likelihood estimation of a log-concave density and its distribution function: basic properties and uniform consistency
• Title not available (Why is that?)
• Multivariate extensions of isotonic regression and total variation denoising via entire monotonicity and Hardy-Krause variation
• Title not available (Why is that?)
• The asymptotic behavior of monotone regression estimates
• A general projection framework for constrained smoothing.
• Two estimators of the mean of a counting process with panel count data.
• Likelihood ratio tests for monotone functions.
• Estimation of a \(k\)-monotone density: limit distribution theory and the spline connection
• Title not available (Why is that?)
• Title not available (Why is that?)
• Bandwidth choice for nonparametric regression
• Representation theorem for convex nonparametric least squares
• Point Estimates of Ordinates of Concave Functions
• Inconsistency of bootstrap: the Grenander estimator
• Nonparametric estimation under shape constraints. Estimators, algorithms and asymptotics
• Title not available (Why is that?)
• Nonparametric estimation of a convex bathtub-shaped hazard function
• Confidence sets for split points in decision trees
• Title not available (Why is that?)
• Estimation of a convex function: Characterizations and asymptotic theory.
• Global rates of convergence of the MLEs of log-concave and \(s\)-concave densities
• Approximation and estimation of \(s\)-concave densities via Rényi divergences
• Global rates of convergence in log-concave density estimation
• Title not available (Why is that?)
• Quasi-concave density estimation
• Nonparametric estimation of multivariate convex-transformed densities
• Global risk bounds and adaptation in univariate convex regression
• Nonparametric regression under qualitative smoothness assumptions
• Consistency in concave regression
• Concave regression: value-constrained estimation and likelihood ratio-based inference
• A canonical process for estimation of convex functions: the ``invelope of integrated Brownian motion \(+t^ 4\).
• Risk bounds in isotonic regression
• On risk bounds in isotonic and other shape restricted regression problems
• Likelihood based inference for monotone response models
• Consistency of multidimensional convex regression
• On Difference-Based Variance Estimation in Nonparametric Regression When the Covariate is High Dimensional
• Sharp oracle inequalities for least squares estimators in shape restricted regression
• Estimation for Distributions with Monotone Failure Rate
• Isotonic estimation and rates of convergence in Wicksell's problem
• Change-point in stochastic design regression and the bootstrap
• Beyond Parametrics in Interdisciplinary Research: Festschrift in Honor of Professor Pranab K. Sen
• A pseudolikelihood method for analyzing interval censored data
• Nonparametric least squares estimation of a multivariate convex regression function
• On estimation of noise variance in two-dimensional signal processing
• Title not available (Why is that?)
• On univariate convex regression
• Adaptive risk bounds in unimodal regression
• Isotonic regression in multi-dimensional spaces and graphs
• Isotonic regression in general dimensions
• Nonparametric confidence intervals for monotone functions
• On matrix estimation under monotonicity constraints
• Adaptation in log-concave density estimation
• High-dimensional nonparametric density estimation via symmetry and shape constraints
• Adaptation in multivariate log-concave density estimation
• Limit distribution theory for block estimators in multiple isotonic regression
• On Integrated L¹ Convergence Rate of an Isotonic Regression Estimator for Multivariate Observations

Cited In (5)

• Coverage of credible intervals in Bayesian multivariate isotonic regression
• Nonparametric confidence intervals for monotone functions
• Stepwise Confidence Intervals for Monotone Dose–Response Studies
• Nonparametric inference under a monotone hazard ratio order
• Inference for Local Parameters in Convexity Constrained Models

Uses Software

• scar
• GitHub

This page was built for publication: Confidence intervals for multiple isotonic regression and other monotone models

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2054474)