Bootstrap testing of the rank of a matrix via least-squared constrained estimation

DOI10.1080/01621459.2013.847841MaRDI QIDQ4975340zbMATH OpenOpenAlexFDO

Authors François Portier, Bernard Delyon

Publication date 4 August 2017

Published in Journal of the American Statistical Association (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/1301.0768

zbMATH Keywords

dimension reductionhypothesis testingrank estimation

Mathematics Subject Classification ID

Parametric hypothesis testing (62F03) Bootstrap, jackknife and other resampling methods (62F40)

Abstract: In order to test if an unknown matrix has a given rank (null hypothesis), we consider the family of statistics that are minimum squared distances between an estimator and the manifold of fixed-rank matrix. Under the null hypothesis, every statistic of this family converges to a weighted chi-squared distribution. In this paper, we introduce the constrained bootstrap to build bootstrap estimate of the law under the null hypothesis of such statistics. As a result, the constrained bootstrap is employed to estimate the quantile for testing the rank. We provide the consistency of the procedure and the simulations shed light one the accuracy of the constrained bootstrap with respect to the traditional asymptotic comparison. More generally, the results are extended to test if an unknown parameter belongs to a sub-manifold locally smooth. Finally, the constrained bootstrap is easy to compute, it handles a large family of tests and it works under mild assumptions.

Recommendations

• TESTS OF RANK
• A unifying theory of tests of rank
• Improved inference on the rank of a matrix
• Bootstrap tests and confidence regions for functions of a covariance matrix
• Statistical Tests and Estimators of the Rank of a Matrix and Their Applications in Econometric Modelling

Cites work

• A bootstrap method for assessing the dimension of a general regression problem
• An F Approximation to the Distribution of a Linear Combination of Chi-squared Variables.
• Asymptotic Statistics
• Corrections to test statistics in principal Hessian directions
• Dimension estimation in sufficient dimension reduction: a unifying approach
• Extending Sliced Inverse Regression
• Generalized reduced rank tests using the singular value decomposition
• scientific article; zbMATH DE number 125649 (Why is no real title available?)
• scientific article; zbMATH DE number 3517666 (Why is no real title available?)
• Inference on low-rank data matrices with applications to microarray data
• Inferring the rank of a matrix
• Intentionally Biased Bootstrap Methods
• On Directional Regression for Dimension Reduction
• Optimal transformation: a new approach for covering the central subspace
• Sufficient Dimension Reduction via Inverse Regression
• The asymptotic distribution of singular values with applications to canonical correlations and correspondence analysis
• The bootstrap and Edgeworth expansion
• The estimating function bootstrap
• Using the Bootstrap to Select One of a New Class of Dimension Reduction Methods

Cited in

(6)

• Granger causality from quantized measurements
• Identification of structural vector autoregressions through higher unconditional moments
• An empirical process view of inverse regression
• Bootstrap-based regularization for low-rank matrix estimation
• Tail inverse regression: dimension reduction for prediction of extremes
• Interpretable sparse SIR for functional data

This page was built for publication: Bootstrap testing of the rank of a matrix via least-squared constrained estimation

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4975340)