Optimal rates of convergence for noisy sparse phase retrieval via thresholded Wirtinger flow

DOI10.1214/16-AOS1443MaRDI QIDQ342686zbMATH OpenOpenAlexFDO

Authors Xiaodong Li, Zongming Ma, T. Tony Cai

Publication date 18 November 2016

Published in The Annals of Statistics (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/1506.03382

zbMATH Keywords

sparse recoveryminimax ratephase retrievaliterative adaptive thresholdingnon-convex empirical riskthresholded gradient method

Mathematics Subject Classification ID

Minimax procedures in statistical decision theory (62C20) Applications of statistics to physics (62P35)

Abstract: This paper considers the noisy sparse phase retrieval problem: recovering a sparse signal $xinmathbb{R}^p$ from noisy quadratic measurements $y_j=({a_j}^{\prime }x{)}^2+epsilo{n}_j$, $j=1,ldots,m$, with independent sub-exponential noise $epsilo{n}_j$. The goals are to understand the effect of the sparsity of $x$ on the estimation precision and to construct a computationally feasible estimator to achieve the optimal rates. Inspired by the Wirtinger Flow [12] proposed for noiseless and non-sparse phase retrieval, a novel thresholded gradient descent algorithm is proposed and it is shown to adaptively achieve the minimax optimal rates of convergence over a wide range of sparsity levels when the $a_j$'s are independent standard Gaussian random vectors, provided that the sample size is sufficiently large compared to the sparsity of $x$.

Recommendations

• Phase retrieval via Wirtinger flow: theory and algorithms
• Phase retrieval via sparse Wirtinger flow
• Sparse signal recovery from phaseless measurements via hard thresholding pursuit
• Phase retrieval for sparse signals
• \(L_{1/2}\)-regularized least absolute deviation method for sparse phase retrieval

Cited in

(34)

• The numerics of phase retrieval
• Gradient descent with random initialization: fast global convergence for nonconvex phase retrieval
• Compressive phase retrieval: Optimal sample complexity with deep generative priors
• First-order methods almost always avoid strict saddle points
• ISLET: fast and optimal low-rank tensor regression via importance sketching
• Rate-optimal perturbation bounds for singular subspaces with applications to high-dimensional statistics
• Robust amplitude method with \(L_{1/2}\)-regularization for compressive phase retrieval
• Robust High-Dimensional Regression with Coefficient Thresholding and Its Application to Imaging Data Analysis
• Implicit regularization in nonconvex statistical estimation: gradient descent converges linearly for phase retrieval, matrix completion, and blind deconvolution
• Cardinality minimization, constraints, and regularization: a survey
• Inertial proximal ADMM for separable multi-block convex optimizations and compressive affine phase retrieval
• An optimal statistical and computational framework for generalized tensor estimation
• Toward a mathematical theory of the crystallographic phase retrieval problem
• Phase retrieval from Fourier measurements with masks
• Fundamental limits of weak recovery with applications to phase retrieval
• Phase retrieval with PhaseLift algorithm
• A geometric analysis of phase retrieval
• Provable sample-efficient sparse phase retrieval initialized by truncated power method
• Constructing confidence intervals for the signals in sparse phase retrieval
• Nearly minimax-optimal rates for noisy sparse phase retrieval via early-stopped mirror descent
• An Efficient and Robust Scalar Auxialiary Variable Based Algorithm for Discrete Gradient Systems Arising from Optimizations
• Performance bounds of the intensity-based estimators for noisy phase retrieval
• On DC based methods for phase retrieval
• Sparse signal recovery from phaseless measurements via hard thresholding pursuit
• Misspecified nonconvex statistical optimization for sparse phase retrieval
• Phase retrieval using alternating minimization in a batch setting
• Model-free nonconvex matrix completion: local minima analysis and applications in memory-efficient kernel PCA
• \(L_{1/2}\)-regularized least absolute deviation method for sparse phase retrieval
• The recovery of complex sparse signals from few phaseless measurements
• Truncated amplitude flow with coded diffraction patterns
• Variational phase retrieval with globally convergent preconditioned proximal algorithm
• Partially Observed Dynamic Tensor Response Regression
• Phase retrieval via sparse Wirtinger flow
• Proof methods for robust low-rank matrix recovery

This page was built for publication: Optimal rates of convergence for noisy sparse phase retrieval via thresholded Wirtinger flow

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q342686)