Estimation of low-rank covariance function

DOI10.1016/J.SPA.2016.04.006MaRDI QIDQ335675zbMATH OpenOpenAlexFDO

Authors Alexandre B. Tsybakov, Vladimir Koltchinskii, K. Lounici

Publication date 2 November 2016

Published in Stochastic Processes and their Applications (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/1504.03009

zbMATH Keywords

adaptationGaussian processnuclear normempirical risk minimizationlow rank covariance functionminimax lower bounds

Mathematics Subject Classification ID

Gaussian processes (60G15) Non-Markovian processes: estimation (62M09)

Abstract: We consider the problem of estimating a low rank covariance function $K(t,u)$ of a Gaussian process $S(t),tin[0,1]$ based on $n$ i.i.d. copies of $S$ observed in a white noise. We suggest a new estimation procedure adapting simultaneously to the low rank structure and the smoothness of the covariance function. The new procedure is based on nuclear norm penalization and exhibits superior performances as compared to the sample covariance function by a polynomial factor in the sample size $n$. Other results include a minimax lower bound for estimation of low-rank covariance functions showing that our procedure is optimal as well as a scheme to estimate the unknown noise variance of the Gaussian process.

Recommendations

• Nonparametric estimation of covariance functions by model selection
• On the nonparametric estimation of covariance functions
• scientific article; zbMATH DE number 1072474
• Low-Rank Covariance Function Estimation for Multidimensional Functional Data
• scientific article; zbMATH DE number 224222

Cites work

• scientific article; zbMATH DE number 3834852 (Why is no real title available?)
• scientific article; zbMATH DE number 1302661 (Why is no real title available?)
• scientific article; zbMATH DE number 739280 (Why is no real title available?)
• Characteristic Classes. (AM-76)
• Functional Data Analysis for Sparse Longitudinal Data
• Hanson-Wright inequality and sub-Gaussian concentration
• High-dimensional covariance matrix estimation with missing observations
• Introduction to nonparametric estimation
• Low rank estimation of smooth kernels on graphs
• Nonparametric estimation of covariance functions by model selection
• Nuclear-norm penalization and optimal rates for noisy low-rank matrix completion
• On the sample covariance matrix estimator of reduced effective rank population matrices, with applications to fPCA
• Properties of principal component methods for functional and longitudinal data analysis

Cited in

(6)

• Robust Covariance Matrix Estimation in Heterogeneous Low Rank Context
• Low-rank multi-parametric covariance identification
• Joint non-parametric estimation of mean and auto-covariances for Gaussian processes
• Optimal cleaning for singular values of cross-covariance matrices
• Low-Rank Covariance Function Estimation for Multidimensional Functional Data
• Degrees of freedom in low rank matrix estimation

This page was built for publication: Estimation of low-rank covariance function

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q335675)