Estimation of smooth functionals in normal models: bias reduction and asymptotic efficiency

DOI10.1214/20-AOS2047MaRDI QIDQ2054520zbMATH OpenOpenAlexFDO

Authors M. Zhilova, Vladimir Koltchinskii

Publication date 3 December 2021

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

Full work available at URL https://arxiv.org/abs/1912.08877, https://projecteuclid.org/journals/annals-of-statistics/volume-49/issue-5/Estimation-of-smooth-functionals-in-normal-models-Bias-reduction/10.1214/20-AOS2047.full

zbMATH Keywords

bias reductionbootstrap chainconcentrationefficiencyrandom homotopysmooth functionals

Mathematics Subject Classification ID

Random matrices (probabilistic aspects) (60B20) Asymptotic properties of nonparametric inference (62G20) Estimation in multivariate analysis (62H12) Factor analysis and principal components; correspondence analysis (62H25)

Abstract: Let $X_1,dots,X_n$ be i.i.d. random variables sampled from a normal distribution $N(mu,Sigma)$ in ${mathbbR}^d$ with unknown parameter $heta=(mu,Sigma)inTheta:={mathbbR}^{d}imes{mathcalC}_{+}^d,$ where ${mathcalC}_{+}^d$ is the cone of positively definite covariance operators in ${mathbbR}^d.$ Given a smooth functional $f:Thetamapsto{mathbbR}^1,$ the goal is to estimate $f(heta)$ based on $X_1,dots,X_n.$ Let ${\displaystyle }$ Theta(a;d):={mathbb R}^d imes Bigl{Sigmain {mathcal C}_+^d: sigma(Sigma)subset [1/a, a]Bigr}, ageq 1, ${\displaystyle }$ where $sigma(Sigma)$ is the spectrum of covariance $Sigma.$ Let $hatheta:=(hatmu,hatSigma),$ where $hatmu$ is the sample mean and $hatSigma$ is the sample covariance, based on the observations $X_1,dots,X_n.$ For an arbitrary functional $fin{C}^s(Theta),$ $s=k+1+ho,kgeq0,hoin(0,1],$ we define a functional $f_k:ThetamapstomathbbR$ such that �egin{align*} & sup_{ hetain Theta(a;d)}|f_k(hat heta)-f( heta)|_{L_2({mathbb P}_{ heta})} lesssim_{s, �eta} |f|_{C^{s}(Theta)} �iggr[�iggl(frac{a}{sqrt{n}} �igvee a^{�eta s}�iggl(sqrt{frac{d}{n}}�iggr)^{s} �iggr)wedge 1�iggr], end{align*} where $$ for $k=0$ and $$ is arbitrary for $kgeq1.$ This error rate is minimax optimal and similar bounds hold for more general loss functions. If $d=d_{n}leq{n}^{alpha}$ for some $alphain(0,1)$ and $sgeqfrac11-alpha,$ the rate becomes $O(n^{-1/2}).$ Moreover, for $s>frac11-alpha,$ the estimators $f_k(hatheta)$ is shown to be asymptotically efficient. The crucial part of the construction of estimator $f_k(hatheta)$ is a bias reduction method studied in the paper for more general statistical models than normal.

Recommendations

• Asymptotically efficient estimation of smooth functionals of covariance operators
• Asymptotic efficiency in high-dimensional covariance estimation
• Efficient estimation of smooth functionals in Gaussian shift models
• Efficient estimation of linear functionals of principal components
• scientific article; zbMATH DE number 850368

Cites work

• Applications of the van Trees inequality: A Bayesian Cramér-Rao bound
• Asymptotic efficiency in high-dimensional covariance estimation
• Asymptotic normality of quadratic estimators
• Asymptotically efficient estimation of smooth functionals of covariance operators
• Asymptotics and concentration bounds for bilinear forms of spectral projectors of sample covariance
• Bias Correction With Jackknife, Bootstrap, and Taylor Series
• Combinatorics of partial derivatives
• Concentration inequalities and moment bounds for sample covariance operators
• Concentration inequalities for non-Lipschitz functions with bounded derivatives of higher order
• Efficient estimation of integral functionals of a density
• Efficient estimation of linear functionals of principal components
• Efficient estimation of smooth functionals in Gaussian shift models
• Estimation of integral functionals of a density
• Estimation of smooth functionals of location parameter in Gaussian and Poincaré random shift models.
• Geometrizing rates of convergence. II
• Higher order influence functions and minimax estimation of nonlinear functionals
• Higher order tangent spaces and influence functions
• scientific article; zbMATH DE number 3980111 (Why is no real title available?)
• scientific article; zbMATH DE number 4107941 (Why is no real title available?)
• scientific article; zbMATH DE number 3656968 (Why is no real title available?)
• scientific article; zbMATH DE number 490141 (Why is no real title available?)
• scientific article; zbMATH DE number 1522808 (Why is no real title available?)
• Introduction to General Statistical Analysis
• Minimax estimation of linear and quadratic functionals on sparsity classes
• Minimax quadratic estimation of a quadratic functional
• Nonquadratic estimators of a quadratic functional
• Normal approximation and concentration of spectral projectors of sample covariance
• On adaptive estimation of linear functionals
• On estimation of L_r-norms in Gaussian white noise models
• On estimation of the \(L_r\) norm of a regression function
• On Necessary Conditions for Efficient Estimation of Functionals of a Nonparametric Signal in White Noise
• On the Efficiency of a Class of Non-Parametric Estimates
• Operator Lipschitz functions
• Optimal Transport
• Perturbation of linear forms of singular vectors under Gaussian noise
• Representation of Markov chains by random maps: existence and regularity conditions
• Sharp adaptive estimation of quadratic functionals
• Some Problems on Nonparametric Estimation in Gaussian White Noise
• Statistical analysis of observations of increasing dimension. Transl. from the Russian
• Theory of function spaces

Cited in

(11)

• Efficient estimation of smooth functionals in Gaussian shift models
• Estimation of smooth functionals in high-dimensional models: bootstrap chains and Gaussian approximation
• Asymptotically efficient estimation of smooth functionals of covariance operators
• Asymptotic efficiency in high-dimensional covariance estimation
• Functional estimation in log-concave location families
• Sharp concentration of simple random tensors
• On the estimation of Gaussian moment tensors
• Concentration and consistency of sample covariance matrix functionals in sub-Gaussian models
• Estimation of smooth functionals of covariance operators: jackknife bias reduction and bounds in terms of effective rank
• Universal bias reduction in estimation of smooth additive function in high dimensions
• Möbius inversion and the iterated bootstrap

This page was built for publication: Estimation of smooth functionals in normal models: bias reduction and asymptotic efficiency

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2054520)