UT logo

SVP: Singular Value Projection

About | Downloads | Documentation | Contact

Latest version: 1.1 (Source code only)
                       1.1 (pre-compiled 32 bits, Ubuntoo)
                       1.1 (pre-compiled 64 bits, Ubuntoo)

About SVP

SVP is a Matlab implementation of the Singular Value Projection Method. SVP solves the problem of low-rank matrix completion by using a variant of the classical projected gradient method. It uses PROPACK (written by Rasmus Munk Larsen) for computing SVD at each step for projection. We have modified the wrapper files lansvd and lanbpro file of PROPACK to suit our needs.

This software is currently being actively maintained; please check back often for updates. When using this code, please cite SVP and the relevant paper.

@article{MekaJD09,
  author    = {Raghu Meka and
               Prateek Jain and
               Inderjit S. Dhillon},
  title     = {Guaranteed Rank Minimization via Singular Value Projection},
  journal   = {CoRR},
  volume    = {abs/0909.5457},
  year      = {2009},
  ee        = {http://arxiv.org/abs/0909.5457}
}

@manual{svp,
  title        = {Singular Value Projection},
  author       = {Prateek Jain and Raghu Meka and Inderjit S. Dhillon},
  organization = {UT, Austin},
  address      = {http://www.cs.utexas.edu/users/pjain/svp/},
}

Downloads

Source code

Download the latest version (1.1, released 2010-03-11): 1.1 (Source code only)
                                                                                      svp-1.1_32.tar.gz (pre-compiled 32 bits, Ubuntoo)
                                                                                       svp-1.1_64.tar.gz (pre-compiled 64 bits, Ubuntoo)


Documentation

README: How to install and use SVP. Also, includes answers to some FAQs.
mexopts.sh

Contact

If you have any questions, suggestions, or bug reports about this implementation, please contact Prateek Jain (pjain at cs dot utexas dot edu) and Raghu Meka (raghu at cs dot utexas dot edu).