The Matrix Factorization Jungle Page.

It surprises me to no end. The new page on Matrix Factorization has been seen on five continents one day after I made it live. The idea for the page stems from the relative scattered information on these new factorizations that have appeared under the impulse of fields like compressive sensing, software like CVX and techniques like the interior point methods. While before say 2006, one could literally drown under different kinds of PCA techniques, we should expect a similar fate for these new factorization techniques. The sources used to set up this list include :

• Stephen Becker's page,
• Raghunandan H. Keshavan's page,
• Nuclear Norm and Matrix Recovery through SDP by Christoph Helmberg, 
• Arvind Ganesh's Low-Rank Matrix Recovery and Completion via Convex Optimization 

who provide more in-depth additional information. Additional codes were featured also on Nuit Blanche.

The idea is that the page should be a place that links to pages that are dedicated to an implementation of a new algorithm that supports the finding of the attendant paper/publication.

Most of the algorithms listed so far generally rely on using the nuclear norm as a proxy to the rank functional even though it may not be optimal. CVX ( Michael Grant and Stephen Boyd) consistently allows one to explore other proxies for the rank functional such as the log-det as found by Maryam Fazell, Haitham Hindi, Stephen Boyd. 

Currently the page needs to be updated with some information about authors. Any additional insights are obviously welcome.

Liked this entry ? subscribe to the Nuit Blanche feed, there's more where that came from