On Convergence of the Alternating Projection Method for Matrix Completion and Sparse Recovery Problems
arXiv:1711.02151
Abstract
In this paper, we study the convergence of Alternating Projection (AP) algorithm for the matrix completion and compressed sensing problems. We also present computational evidence for the excellent performance of the algorithm. Also, in the last section, we prove using algebraic-geometric techniques that, fixing the known positions, if a rank r matrix can be completed in finitely many ways using one given set of known entries, then, for almost all set of known entries, the matrix can be only completed into a rank r matrix in finitely many ways.
arXiv comment29
References in corpus (7)
- Guaranteed Minimum-Rank Solutions of Linear Matrix Equations via Nuclear Norm Minimization
- Message Passing Algorithms for Compressed Sensing
- An overview of low-rank matrix recovery from incomplete observations
- A General Iterative Shrinkage and Thresholding Algorithm for Non-convex Regularized Optimization Problems
- Augmented L1 and Nuclear-Norm Models with a Globally Linearly Convergent Algorithm
- New Null Space Results and Recovery Thresholds for Matrix Rank Minimization
- Guarantees of Riemannian Optimization for Low Rank Matrix Completion
Cited by in corpus (5)
- CPGD: Cadzow Plug-and-Play Gradient Descent for Generalised FRI
- Provable Subspace Tracking from Missing Data and Matrix Completion
- On Asymptotic Linear Convergence Rate of Iterative Hard Thresholding for Matrix Completion
- Unbiased Subdata Selection for Fair Classification: A Unified Framework and Scalable Algorithms
- Robust Online Video Super-Resolution Using an Efficient Alternating Projections Scheme