Tuesday, January 08, 2013

A Probabilistic Approach to Robust Matrix Factorization - implementation -

Naiyan Wang sent me the following:

Dear Igor,

First, thanks a lot for maintaining such a wonderful blog. It is a great source to hear latest news in the related area. Next, I would like to introduce you our paper on robust matrix factorization: "A Probabilistic Approach to Robust Matrix Factorization", which is published in ECCV12'. We have set up the project page here: http://winsty.net/prmf.html . You could find the codes and demos there. Could you please kindly add it to the matrix factorization jungle?
Thanks
Naiyan

Sure Naiyan ! so it looks like we have a good contender to GoDec, the solver Cable and I use to speed up the CAI series: This is very interesting:






Matrix factorization underlies a large variety of computer vision applications. It is a particularly challenging problem for large-scale applications and when there exist outliers and missing data. In this paper, we propose a novel probabilistic model called Probabilistic Robust Matrix Factorization (PRMF) to solve this problem. In particular, PRMF is formulated with a Laplace error and a Gaussian prior which correspond to an l1 loss and an l2 regularizer, respectively. For model learning, we devise a parallelizable expectation-maximization (EM) algorithm which can potentially be applied to large-scale applications. We also propose an online extension of the algorithm for sequential data to offer further scalability. Experiments conducted on both synthetic data and some practical computer vision applications show that PRMF is comparable to other state-of-the-art robust matrix factorization methods in terms of accuracy and outperforms them particularly for large data matrices.

No comments:

Printfriendly