Implicit Regularization in Nonconvex Statistical Estimation: Gradient Descent Converges Linearly for Phase Retrieval, Matrix Completion, and Blind Deconvolution
Published 2019 View Full Article
- Home
- Publications
- Publication Search
- Publication Details
Title
Implicit Regularization in Nonconvex Statistical Estimation: Gradient Descent Converges Linearly for Phase Retrieval, Matrix Completion, and Blind Deconvolution
Authors
Keywords
Nonconvex optimization, Gradient descent, Leave-one-out analysis, Phase retrieval, Matrix completion, Blind deconvolution, 90C26
Journal
FOUNDATIONS OF COMPUTATIONAL MATHEMATICS
Volume -, Issue -, Pages -
Publisher
Springer Science and Business Media LLC
Online
2019-08-06
DOI
10.1007/s10208-019-09429-9
References
Ask authors/readers for more resources
Related references
Note: Only part of the references are listed.- Gradient descent with random initialization: fast global convergence for nonconvex phase retrieval
- (2019) Yuxin Chen et al. MATHEMATICAL PROGRAMMING
- BranchHull: Convex bilinear inversion from the entrywise product of signals with known signs
- (2019) Alireza Aghasi et al. APPLIED AND COMPUTATIONAL HARMONIC ANALYSIS
- Structured Signal Recovery From Quadratic Measurements: Breaking Sample Complexity Barriers via Nonconvex Optimization
- (2019) Mahdi Soltanolkotabi IEEE TRANSACTIONS ON INFORMATION THEORY
- Spectral method and regularized MLE are both optimal for top-$K$ ranking
- (2019) Yuxin Chen et al. ANNALS OF STATISTICS
- Nonconvex Optimization Meets Low-Rank Matrix Factorization: An Overview
- (2019) Yuejie Chi et al. IEEE TRANSACTIONS ON SIGNAL PROCESSING
- Rapid, robust, and reliable blind deconvolution via nonconvex optimization
- (2018) Xiaodong Li et al. APPLIED AND COMPUTATIONAL HARMONIC ANALYSIS
- Fast and Guaranteed Blind Multichannel Deconvolution Under a Bilinear System Model
- (2018) Kiryung Lee et al. IEEE TRANSACTIONS ON INFORMATION THEORY
- PhaseMax: Convex Phase Retrieval via Basis Pursuit
- (2018) Tom Goldstein et al. IEEE TRANSACTIONS ON INFORMATION THEORY
- Theoretical insights into the optimization landscape of over-parameterized shallow neural networks
- (2018) Mahdi Soltanolkotabi et al. IEEE TRANSACTIONS ON INFORMATION THEORY
- Sparse Phase Retrieval via Truncated Amplitude Flow
- (2018) Gang Wang et al. IEEE TRANSACTIONS ON SIGNAL PROCESSING
- Near-Optimal Bounds for Phase Synchronization
- (2018) Yiqiao Zhong et al. SIAM JOURNAL ON OPTIMIZATION
- The landscape of empirical risk for nonconvex losses
- (2018) Song Mei et al. ANNALS OF STATISTICS
- Debiasing the lasso: Optimal sample size for Gaussian designs
- (2018) Adel Javanmard et al. ANNALS OF STATISTICS
- The Projected Power Method: An Efficient Algorithm for Joint Alignment from Pairwise Differences
- (2018) Yuxin Chen et al. COMMUNICATIONS ON PURE AND APPLIED MATHEMATICS
- Fast Rank-One Alternating Minimization Algorithm for Phase Retrieval
- (2018) Jian-Feng Cai et al. JOURNAL OF SCIENTIFIC COMPUTING
- Blind Recovery of Sparse Signals From Subsampled Convolution
- (2017) Kiryung Lee et al. IEEE TRANSACTIONS ON INFORMATION THEORY
- Complete Dictionary Recovery Over the Sphere I: Overview and the Geometric Picture
- (2017) Ju Sun et al. IEEE TRANSACTIONS ON INFORMATION THEORY
- Optimal rates of convergence for noisy sparse phase retrieval via thresholded Wirtinger flow
- (2016) T. Tony Cai et al. ANNALS OF STATISTICS
- Low rank matrix completion by alternating steepest descent methods
- (2016) Jared Tanner et al. APPLIED AND COMPUTATIONAL HARMONIC ANALYSIS
- Solving Random Quadratic Systems of Equations Is Nearly as Easy as Solving Linear Systems
- (2016) Yuxin Chen et al. COMMUNICATIONS ON PURE AND APPLIED MATHEMATICS
- Guaranteed Blind Sparse Spikes Deconvolution via Lifting and Convex Optimization
- (2016) Yuejie Chi IEEE Journal of Selected Topics in Signal Processing
- An Overview of Low-Rank Matrix Recovery From Incomplete Observations
- (2016) Mark A. Davenport et al. IEEE Journal of Selected Topics in Signal Processing
- Blind Deconvolution From Multiple Sparse Inputs
- (2016) Liming Wang et al. IEEE SIGNAL PROCESSING LETTERS
- Kaczmarz Method for Solving Quadratic Equations
- (2016) Yuejie Chi et al. IEEE SIGNAL PROCESSING LETTERS
- Guaranteed Matrix Completion via Non-Convex Factorization
- (2016) Ruoyu Sun et al. IEEE TRANSACTIONS ON INFORMATION THEORY
- The Local Convexity of Solving Systems of Quadratic Equations
- (2016) Sujay Sanghavi et al. Results in Mathematics
- Guarantees of Riemannian Optimization for Low Rank Matrix Recovery
- (2016) Ke Wei et al. SIAM JOURNAL ON MATRIX ANALYSIS AND APPLICATIONS
- ROP: Matrix recovery via rank-one projections
- (2015) T. Tony Cai et al. ANNALS OF STATISTICS
- A useful variant of the Davis–Kahan theorem for statisticians
- (2015) Y. Yu et al. BIOMETRIKA
- Incoherence-Optimal Matrix Completion
- (2015) Yudong Chen IEEE TRANSACTIONS ON INFORMATION THEORY
- Exact and Stable Covariance Estimation From Quadratic Sampling via Convex Programming
- (2015) Yuxin Chen et al. IEEE TRANSACTIONS ON INFORMATION THEORY
- Phase Retrieval via Wirtinger Flow: Theory and Algorithms
- (2015) Emmanuel J. Candes et al. IEEE TRANSACTIONS ON INFORMATION THEORY
- Compressive Phase Retrieval via Generalized Approximate Message Passing
- (2015) Philip Schniter et al. IEEE TRANSACTIONS ON SIGNAL PROCESSING
- Self-calibration and biconvex compressive sensing
- (2015) Shuyang Ling et al. INVERSE PROBLEMS
- Solving systems of phaseless equations via Kaczmarz methods: a proof of concept study
- (2015) Ke Wei INVERSE PROBLEMS
- Blind Deconvolution Using Convex Programming
- (2014) Ali Ahmed et al. IEEE TRANSACTIONS ON INFORMATION THEORY
- GESPAR: Efficient Phase Retrieval of Sparse Signals
- (2014) Yoav Shechtman et al. IEEE TRANSACTIONS ON SIGNAL PROCESSING
- Hanson-Wright inequality and sub-gaussian concentration
- (2013) Mark Rudelson et al. ELECTRONIC COMMUNICATIONS IN PROBABILITY
- Solving Quadratic Equations via PhaseLift When There Are About as Many Equations as Unknowns
- (2013) Emmanuel J. Candès et al. FOUNDATIONS OF COMPUTATIONAL MATHEMATICS
- On robust regression with high-dimensional predictors
- (2013) Noureddine El Karoui et al. PROCEEDINGS OF THE NATIONAL ACADEMY OF SCIENCES OF THE UNITED STATES OF AMERICA
- Phase Retrieval via Matrix Completion
- (2013) Emmanuel J. Candès et al. SIAM Journal on Imaging Sciences
- PhaseLift: Exact and Stable Signal Recovery from Magnitude Measurements via Convex Programming
- (2012) Emmanuel J. Candès et al. COMMUNICATIONS ON PURE AND APPLIED MATHEMATICS
- A tail inequality for quadratic forms of subgaussian random vectors
- (2012) Daniel Hsu et al. ELECTRONIC COMMUNICATIONS IN PROBABILITY
- Nuclear-norm penalization and optimal rates for noisy low-rank matrix completion
- (2011) Vladimir Koltchinskii et al. ANNALS OF STATISTICS
- Recovering Low-Rank Matrices From Few Coefficients in Any Basis
- (2011) David Gross IEEE TRANSACTIONS ON INFORMATION THEORY
- A Probabilistic and RIPless Theory of Compressed Sensing
- (2011) Emmanuel J. Candes et al. IEEE TRANSACTIONS ON INFORMATION THEORY
- Robust principal component analysis?
- (2011) Emmanuel J. Candès et al. JOURNAL OF THE ACM
- Rank-Sparsity Incoherence for Matrix Decomposition
- (2011) Venkat Chandrasekaran et al. SIAM JOURNAL ON OPTIMIZATION
- Matrix Completion From a Few Entries
- (2010) Raghunandan H. Keshavan et al. IEEE TRANSACTIONS ON INFORMATION THEORY
- The Power of Convex Relaxation: Near-Optimal Matrix Completion
- (2010) Emmanuel J. Candes et al. IEEE TRANSACTIONS ON INFORMATION THEORY
- ADMiRA: Atomic Decomposition for Minimum Rank Approximation
- (2010) Kiryung Lee et al. IEEE TRANSACTIONS ON INFORMATION THEORY
- Guaranteed Minimum-Rank Solutions of Linear Matrix Equations via Nuclear Norm Minimization
- (2010) Benjamin Recht et al. SIAM REVIEW
- Exact Matrix Completion via Convex Optimization
- (2009) Emmanuel J. Candès et al. FOUNDATIONS OF COMPUTATIONAL MATHEMATICS
Publish scientific posters with Peeref
Peeref publishes scientific posters from all research disciplines. Our Diamond Open Access policy means free access to content and no publication fees for authors.
Learn MoreAsk a Question. Answer a Question.
Quickly pose questions to the entire community. Debate answers and get clarity on the most important issues facing researchers.
Get Started