Analysis of Spectral Methods for Phase Retrieval With Random Orthogonal Matrices

DOI10.1109/TIT.2020.2981910MaRDI QIDQ5124496zbMATH OpenOpenAlexFDO

Authors Rishabh Dudeja, Milad Bakhshizadeh, Junjie Ma, Arian Maleki

Publication date 29 September 2020

Published in IEEE Transactions on Information Theory (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/1903.02676

Mathematics Subject Classification ID

Orthogonal matrices (15B10) Random matrices (algebraic aspects) (15B52) Signal theory (characterization, reconstruction, filtering, etc.) (94A12)

Abstract: Phase retrieval refers to algorithmic methods for recovering a signal from its phaseless measurements. Local search algorithms that work directly on the non-convex formulation of the problem have been very popular recently. Due to the nonconvexity of the problem, the success of these local search algorithms depends heavily on their starting points. The most widely used initialization scheme is the spectral method, in which the leading eigenvector of a data-dependent matrix is used as a starting point. Recently, the performance of the spectral initialization was characterized accurately for measurement matrices with independent and identically distributed entries. This paper aims to obtain the same level of knowledge for isotropically random column-orthogonal matrices, which are substantially better models for practical phase retrieval systems. Towards this goal, we consider the asymptotic setting in which the number of measurements $m$, and the dimension of the signal, $n$, diverge to infinity with $m/n=deltain(1,infty)$, and obtain a simple expression for the overlap between the spectral estimator and the true signal vector.

Cited in

(7)

• Optimal combination of linear and spectral estimators for generalized linear models
• Complex phase retrieval from subgaussian measurements
• Approximate message passing with spectral initialization for generalized linear models*
• Spectrum Truncation Power Iteration for Agnostic Matrix Phase Retrieval
• Spiked multiplicative random matrices and principal components
• Spectral estimators for structured generalized linear models via approximate message passing
• Spectrum-aware debiasing: a modern inference framework with applications to principal components regression

This page was built for publication: Analysis of Spectral Methods for Phase Retrieval With Random Orthogonal Matrices

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q5124496)