Optimal estimation of variance in nonparametric regression with random design

DOI10.1214/20-AOS1944zbMATH Open1460.62053arXiv1902.10822MaRDI QIDQ1996785FDOQ1996785

Authors: Yandi Shen, Chao Gao, Fang Han, Daniela M. Witten 

Publication date: 26 February 2021

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

Abstract: Consider the heteroscedastic nonparametric regression model with random design �egin{align*} Y_i = f(X_i) + V^{1/2}(X_i)varepsilon_i, quad i=1,2,ldots,n, end{align*} with $f(cdot)$ and $V(cdot)$ $alpha$- and $$-H"older smooth, respectively. We show that the minimax rate of estimating $V(cdot)$ under both local and global squared risks is of the order �egin{align*} n^{-frac{8alpha�eta}{4alpha�eta + 2alpha + �eta}} vee n^{-frac{2�eta}{2�eta+1}}, end{align*} where $aveeb:=maxa,b$ for any two real numbers $a,b$. This result extends the fixed design rate $$ derived in Wang et al. [2008] in a non-trivial manner, as indicated by the appearances of both $alpha$ and $$ in the first term. In the special case of constant variance, we show that the minimax rate is $n^{-8alpha/(4alpha+1)}vee{n}^{-1}$ for variance estimation, which further implies the same rate for quadratic functional estimation and thus unifies the minimax rate under the nonparametric regression model with those under the density model and the white noise model. To achieve the minimax rate, we develop a U-statistic-based local polynomial estimator and a lower bound that is constructed over a specified distribution family of randomness designed for both $varepsilo{n}_i$ and $X_i$.

Full work available at URL: https://arxiv.org/abs/1902.10822

Recommendations

• Minimax exact constant in sup-norm for nonparametric regression with random design
• scientific article; zbMATH DE number 947421
• scientific article; zbMATH DE number 4100415
• Rate-optimal nonparametric estimation for random coefficient regression models
• Variance function estimation in multivariate nonparametric regression with fixed design

zbMATH Keywords

nonparametric regressionvariance estimationU-statisticsminimax raterandom design

Mathematics Subject Classification ID

Nonparametric regression and quantile regression (62G08) Asymptotic properties of nonparametric inference (62G20) Minimax procedures in statistical decision theory (62C20) Response surface designs (62K20)

Cites Work

• Adaptive estimation of high-dimensional signal-to-noise ratios
• Efficient estimation of conditional variance functions in stochastic regression
• Local Polynomial Variance-Function Estimation
• Distribution of the Ratio of the Mean Square Successive Difference to the Variance
• Estimation of heteroscedasticity in regression analysis
• Nonparametric estimation of global functionals and a measure of the explanatory power of covariates in regression
• Local linear regression smoothers and their minimax efficiencies
• Anisotropic function estimation using multi-bandwidth Gaussian processes
• An Effective Bandwidth Selector for Local Least Squares Regression
• Asymptotically optimal difference-based estimation of variance in nonparametric regression
• Higher order influence functions and minimax estimation of nonlinear functionals
• Design-adaptive Nonparametric Regression
• Introduction to nonparametric estimation
• Risk bounds for model selection via penalization
• Semiparametric minimax rates
• Efficient estimation of integral functionals of a density
• Adaptive estimation of a quadratic functional by model selection.
• Bandwidth choice for nonparametric regression
• On estimating a density using Hellinger distance and some other strange facts
• Asymptotically Minimax Adaptive Estimation. I: Upper Bounds. Optimally Adaptive Estimates
• Residual variance and residual pattern in nonlinear regression
• Title not available (Why is that?)
• Local polynomial estimators of the volatility function in nonparametric autoregression
• Title not available (Why is that?)
• Title not available (Why is that?)
• Title not available (Why is that?)
• Random rates in anisotropic regression. (With discussion)
• Variance estimation in high-dimensional linear models
• Effect of mean on variance function estimation in nonparametric regression
• Variance estimation in nonparametric regression via the difference sequence method
• On Difference-Based Variance Estimation in Nonparametric Regression When the Covariate is High Dimensional
• Variance function estimation in multivariate nonparametric regression with fixed design
• Title not available (Why is that?)
• On a Problem of Adaptive Estimation in Gaussian White Noise
• Goodness-of-fit tests for high-dimensional Gaussian linear models
• On variance estimation in nonparametric regression
• Estimating the error variance in nonparametric regression by a covariate-matched u-statistic
• Title not available (Why is that?)
• Adaptive variance function estimation in heteroscedastic nonparametric regression
• Nonparametric estimation of quadratic regression functionals
• On optimal adaptive estimation of a quadratic functional
• Minimax quadratic estimation of a quadratic functional
• On the estimation of quadratic functionals
• Optimal adaptive estimation of a quadratic functional
• Variance estimation for high-dimensional regression models
• Nonparametric Estimation of a Regression Function that is Smooth in a Domain in $R^k$
• A Further Remark Concerning the Distribution of the Ratio of the Mean Square Successive Difference to the Variance
• A simple adaptive estimator of the integrated square of a density
• Noise estimation in signal restoration using regularization
• Minimax Rates for Estimating the Variance and its Derivatives in Non–Parametric Regression

Cited In (8)

• The max-MSE's of minimax estimators of variance in nonparametric regression
• Minimax Rates for Estimating the Variance and its Derivatives in Non–Parametric Regression
• Title not available (Why is that?)
• Approximate tolerance intervals for nonparametric regression models
• Optimal variance estimation without estimating the mean function
• Minimax exact constant in sup-norm for nonparametric regression with random design
• Inference for Local Parameters in Convexity Constrained Models
• Wavelet estimation for the nonparametric additive model in random design and long-memory dependent errors

This page was built for publication: Optimal estimation of variance in nonparametric regression with random design

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q1996785)