Corpus ID: 58028755

# Nonconvex Rectangular Matrix Completion via Gradient Descent without $\ell_{2,\infty}$ Regularization

@article{Chen2019NonconvexRM,
title={Nonconvex Rectangular Matrix Completion via Gradient Descent without \$\ell_\{2,\infty\}\$ Regularization},
author={Ji ming Chen and Dekai Liu and Xiaoguang Li},
journal={arXiv: Machine Learning},
year={2019}
}
• Published 2019
• Mathematics, Computer Science
• arXiv: Machine Learning
• The analysis of nonconvex matrix completion has recently attracted much attention in the community of machine learning thanks to its computational convenience. Existing analysis on this problem, however, usually relies on $\ell_{2,\infty}$ projection or regularization that involves unknown model parameters, although they are observed to be unnecessary in numerical simulations, see, e.g., Zheng and Lafferty [2016]. In this paper, we extend the analysis of the vanilla gradient descent for… CONTINUE READING

#### Citations

##### Publications citing this paper.
SHOWING 1-9 OF 9 CITATIONS

## Nonconvex Low-Rank Symmetric Tensor Completion from Noisy Data

• Mathematics, Computer Science
• ArXiv
• 2019

## Bridging Convex and Nonconvex Optimization in Robust PCA: Noise, Outliers, and Missing Data

• Mathematics, Computer Science, Engineering
• ArXiv
• 2020

## Nonconvex Optimization Meets Low-Rank Matrix Factorization: An Overview

• Psychology, Computer Science, Engineering, Mathematics
• IEEE Transactions on Signal Processing
• 2019

## Noisy Matrix Completion: Understanding Statistical Guarantees for Convex Relaxation via Nonconvex Optimization

• Mathematics, Computer Science
• ArXiv
• 2019

## Inference and uncertainty quantification for noisy matrix completion

• Computer Science, Medicine, Mathematics, Engineering
• Proceedings of the National Academy of Sciences
• 2019

## Inference for linear forms of eigenvectors under minimal eigenvalue separation: Asymmetry and heteroscedasticity

• Computer Science, Mathematics
• ArXiv
• 2020

## Statistical Inferences of Linear Forms for Noisy Matrix Completion

• Computer Science, Mathematics
• ArXiv
• 2019
VIEW 3 EXCERPTS
CITES BACKGROUND & METHODS
HIGHLY INFLUENCED

## Accelerating Ill-Conditioned Low-Rank Matrix Estimation via Scaled Gradient Descent

• Computer Science, Mathematics
• ArXiv
• 2020

## Subspace Estimation from Unbalanced and Incomplete Data Matrices: $\ell_{2,\infty}$ Statistical Guarantees.

• Mathematics, Computer Science
• 2019

#### References

##### Publications referenced by this paper.
SHOWING 1-10 OF 55 REFERENCES

## Algorithmic Regularization in Over-parameterized Matrix Sensing and Neural Networks with Quadratic Activations

• Computer Science, Mathematics
• COLT
• 2018
VIEW 3 EXCERPTS

## Convergence Analysis for Rectangular Matrix Completion Using Burer-Monteiro Factorization and Gradient Descent

• Mathematics, Computer Science
• ArXiv
• 2016
VIEW 7 EXCERPTS
HIGHLY INFLUENTIAL

## Gradient descent with random initialization: fast global convergence for nonconvex phase retrieval

• Mathematics, Computer Science
• Math. Program.
• 2019
VIEW 2 EXCERPTS

## Fast Algorithms for Robust PCA via Gradient Descent

• Computer Science, Mathematics
• NIPS
• 2016

## The Global Optimization Geometry of Nonsymmetric Matrix Factorization and Sensing

• Mathematics, Computer Science
• ArXiv
• 2017

## Memory-efficient Kernel PCA via Partial Matrix Sampling and Nonconvex Optimization: a Model-free Analysis of Local Minima

• Computer Science, Mathematics
• ArXiv
• 2017
VIEW 10 EXCERPTS
HIGHLY INFLUENTIAL

## The Leave-one-out Approach for Matrix Completion: Primal and Dual Analysis

• Computer Science, Mathematics
• ArXiv
• 2018
VIEW 3 EXCERPTS

## Guaranteed Matrix Completion via Nonconvex Factorization

• Computer Science
• 2015 IEEE 56th Annual Symposium on Foundations of Computer Science
• 2015

## Rapid, Robust, and Reliable Blind Deconvolution via Nonconvex Optimization

• Mathematics, Computer Science
• ArXiv
• 2016
VIEW 3 EXCERPTS

## Implicit Regularization in Nonconvex Statistical Estimation: Gradient Descent Converges Linearly for Phase Retrieval, Matrix Completion and Blind Deconvolution

• Computer Science, Mathematics
• Found. Comput. Math.
• 2020
VIEW 27 EXCERPTS
HIGHLY INFLUENTIAL