/EnhancedLowRankMatrixApproximation

Low-rank matrix estimation using convex non-convex prior

Primary LanguageMATLABMIT LicenseMIT

The MATLAB code contained in this directory demonstrates the 
Enhanced Low-Rank Matrix Approximation (ELMA) method [1] for 
estimating a low-rank matrix from its noisy observation.
     
This directory contains the following files

1. demo.m
   This file demonstrates the ELMA method and compares it with
   the nuclear norm minimization (NNM).
         
2. demo2.m
   This file plots the penalty functions and their corresponding
   threshold functions. 
       
Contact: Ankit Parekh (ankit.parekh@nyu.edu)
 
[1] Enhanced Low-Rank Matrix Approximation,
     Ankit Parekh and Ivan Selesnick, 
     IEEE Signal Processing Letters, 2016.