Detection of weak signals in high-dimensional complex-valued data

DOI10.1142/S2010326314500014MaRDI QIDQ5416403zbMATH OpenOpenAlexFDO

Authors Alexei Onatski

Publication date 20 May 2014

Published in Random Matrices: Theory and Applications (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/1207.7098

zbMATH Keywords

asymptotic powercontiguityHarish-Chandra/Itzykson-Zuber integralhypergeometric functionpower envelopesignal detectionsphericity testsspiked covariancesub-critical regimetorus scalar product

Mathematics Subject Classification ID

Orthogonal polynomials and functions in several variables expressible in terms of special functions in one variable (33C50) Asymptotic approximations, asymptotic expansions (steepest descent, etc.) (41A60) Asymptotic properties of parametric tests (62F05) Hypothesis testing in multivariate analysis (62H15) Signal theory (characterization, reconstruction, filtering, etc.) (94A12) Detection theory in information and communication theory (94A13)

Abstract: This paper considers the problem of detecting a few signals in high-dimensional complex-valued Gaussian data satisfying Johnstone's (2001) extit{spiked covariance model}. We focus on the difficult case where signals are weak in the sense that the sizes of the corresponding covariance spikes are below the extit{phase transition threshold} studied in Baik et al (2005). We derive a simple analytical expression for the maximal possible asymptotic probability of correct detection holding the asymptotic probability of false detection fixed. To accomplish this derivation, we establish what we believe to be a new formula for the extit{% Harish-Chandra/Itzykson-Zuber (HCIZ) integral} $in{t}_{mathcalU(p)}{e}^{r(AGB{G}^{-1})}dG$, where $A$ has a deficient rank $r<p$. The formula links the HCIZ integral over $mathcalU(p)$ to an HCIZ integral over a potentially much smaller unitary group $mathcalU(r)$. We show that the formula generalizes to the integrals over orthogonal and symplectic groups. In the most general form, it expresses the hypergeometric function ${}_0{F}_0^{(alpha)}$of two $pimesp$ matrix arguments as a repeated contour integral of the hypergeometric function ${}_0{F}_0^{(alpha)}$of two $rimesr$ matrix arguments.

Recommendations

• Signal detection in high dimension: the multispiked case
• Detecting weak signals in high dimensions
• Optimal signal detection in some spiked random matrix models: likelihood ratio tests and linear spectral statistics
• Asymptotics of sample eigenstructure for a large dimensional spiked covariance model
• Extreme eigenvalues of large-dimensional spiked Fisher matrices with application

Cites work

• A Fourier view on the \(R\)-transform and related asymptotics of spherical integrals
• Asymptotic power of sphericity tests for high-dimensional data
• Complex singular Wishart matrices and applications
• COMPUTATIONALLY EFFICIENT RECURSIONS FOR TOP-ORDER INVARIANT POLYNOMIALS WITH APPLICATIONS
• Differential Operators on a Semisimple Lie Algebra
• DISTRIBUTION OF EIGENVALUES FOR SOME SETS OF RANDOM MATRICES
• Distributions of Matrix Variates and Latent Roots Derived from Normal Samples
• Eigenvalues of large sample covariance matrices of spiked population models
• scientific article; zbMATH DE number 739282 (Why is no real title available?)
• MOPS: multivariate orthogonal polynomials (symbolically)
• New scaling of Itzykson-Zuber integrals
• Non-Parametric Detection of the Number of Signals: Hypothesis Testing and Random Matrix Theory
• On some integrals over theU(N) unitary group and their largeNlimit
• On the distribution of the largest eigenvalue in principal components analysis
• Performance of Statistical Tests for Single-Source Detection Using Random Matrix Theory
• Phase transition of the largest eigenvalue for nonnull complex sample covariance matrices
• Random matrix methods for wireless communications.
• Rank 1 real Wishart spiked model
• Replica field theory for deterministic models. II. A non-random spin glass with glassy behaviour
• Sample Eigenvalue Based Detection of High-Dimensional Signals in White Noise Using Relatively Few Samples
• Statistical Analysis Based on a Certain Multivariate Complex Gaussian Distribution (An Introduction)
• Statistical challenges of high-dimensional data
• Testing Statistical Hypotheses
• The Distributions of Random Matrix Theory and their Applications
• The efficient evaluation of the hypergeometric function of a matrix argument
• The largest eigenvalue of real symmetric, Hermitian and Hermitian self-dual random matrix models with rank one external source. I
• The planar approximation. II
• WKB-Expansion of the HarishChandra-Itzykson-Zuber Integral for Arbitrary

Cited in

(7)

• Rapid evaluation of the spectral signal detection threshold and Stieltjes transform
• Hypergeometric functions of matrix arguments and linear statistics of multi-spiked Hermitian matrix models
• Signal detection in high dimension: the multispiked case
• Detecting weak signals in high dimensions
• Signal Detection in Complex Structured Para Normal Noise
• Multiple change-points detection in high dimension
• Spin glass to paramagnetic transition and triple point in spherical SK model

This page was built for publication: Detection of weak signals in high-dimensional complex-valued data

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q5416403)