Fast nonasymptotic testing and support recovery for large sparse Toeplitz covariance matrices

DOI10.1016/J.JMVA.2021.104883MaRDI QIDQ2140845zbMATH OpenOpenAlexWikidataFDO

Authors Nayel Bettache, Cristina Butucea, Marianne Sorba

Publication date 23 May 2022

Published in Journal of Multivariate Analysis (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/2102.06817

zbMATH Keywords

covariance matrixhigh-dimensional vectorshypothesis testingsparsitysupport recoverytime series

Mathematics Subject Classification ID

Asymptotic properties of parametric estimators (62F12) Hypothesis testing in multivariate analysis (62H15)

Abstract: We consider $n$ independent $p$-dimensional Gaussian vectors with covariance matrix having Toeplitz structure. We test that these vectors have independent components against a stationary distribution with sparse Toeplitz covariance matrix, and also select the support of non-zero entries. We assume that the non-zero values can occur in the recent past (time-lag less than $p/2$). We build test procedures that combine a sum and a scan-type procedures, but are computationally fast, and show their non-asymptotic behaviour in both one-sided (only positive correlations) and two-sided alternatives, respectively. We also exhibit a selector of significant lags and bound the Hamming-loss risk of the estimated support. These results can be extended to the case of nearly Toeplitz covariance structure and to sub-Gaussian vectors. Numerical results illustrate the excellent behaviour of both test procedures and support selectors - larger the dimension $p$, faster are the rates.

Recommendations

• Sharp minimax tests for large Toeplitz covariance matrices with repeated observations
• Large sample test criteria on Toeplitz covariance structure
• Adaptive test for large covariance matrices with missing observations
• Sharp minimax tests for large covariance matrices and adaptation
• Test for high dimensional covariance matrices

Cites work

• A tail inequality for quadratic forms of subgaussian random vectors
• Adaptive detection of a signal of growing dimension. I
• Adaptive detection of a signal of growing dimension. II
• Adaptive thresholding for sparse covariance matrix estimation
• An introduction to recent advances in high/infinite dimensional statistics
• Confidence estimation of the covariance function of stationary and locally stationary processes
• Detecting positive correlations in a multivariate sample
• Detection of a sparse submatrix of a high-dimensional noisy matrix
• Detection of correlations
• Hanson-Wright inequality and sub-Gaussian concentration
• High-dimensional statistics. A non-asymptotic viewpoint
• Higher criticism for detecting sparse heterogeneous mixtures.
• On the range of validity of the autoregressive sieve bootstrap
• Optimal hypothesis testing for high dimensional covariance matrices
• Recent advances in functional data analysis and high-dimensional statistics
• Sharp deviation bounds for quadratic forms
• Sharp minimax tests for large covariance matrices and adaptation
• Sharp minimax tests for large Toeplitz covariance matrices with repeated observations
• The role of statistics in the era of big data: a computational scientist' perspective
• Two-Sample Covariance Matrix Testing and Support Recovery in High-Dimensional and Sparse Settings

Cited in

(2)

• Sharp minimax tests for large Toeplitz covariance matrices with repeated observations
• Large sample test criteria on Toeplitz covariance structure

This page was built for publication: Fast nonasymptotic testing and support recovery for large sparse Toeplitz covariance matrices

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2140845)