3 years ago

Efficient Robust Matrix Factorization with Nonconvex Penalties.

Quanming Yao

Robust matrix factorization (RMF) is a fundamental tool with lots of applications. The state-of-art is robust matrix factorization by majorization and minimization (RMF-MM) algorithm. It iteratively constructs and minimizes a novel surrogate function. Besides, it is also the only RMF algorithm with convergence guarantee. However, it can only deal with the convex $\ell_1$-loss and does not utilize sparsity when matrix is sparsely observed. In this paper, we proposed robust matrix factorization by nonconvex penalties (RMF-NP) algorithm addressing these two problems. RMF-NP enables nonconvex penalties as the loss, which makes it more robust to outliers. As the surrogate function from RMF-MM no longer applies, we construct a new one and solve it in its dual. This makes the runtime and memory cost of RMF-NP only depends on nonzero elements. Convergence analysis based on the new surrogate function is also established, which shows RMF-NP is guaranteed to produce a critical point. Finally, experiments on both synthetic and real-world data sets demonstrate the superiority of RMF-NP over existing algorithms in terms of recovery performance and runtime.

Publisher URL: http://arxiv.org/abs/1710.07205

DOI: arXiv:1710.07205v1

You might also like
Discover & Discuss Important Research

Keeping up-to-date with research can feel impossible, with papers being published faster than you'll ever be able to read them. That's where Researcher comes in: we're simplifying discovery and making important discussions happen. With over 19,000 sources, including peer-reviewed journals, preprints, blogs, universities, podcasts and Live events across 10 research areas, you'll never miss what's important to you. It's like social media, but better. Oh, and we should mention - it's free.

Researcher displays publicly available abstracts and doesn’t host any full article content. If the content is open access, we will direct clicks from the abstracts to the publisher website and display the PDF copy on our platform. Clicks to view the full text will be directed to the publisher website, where only users with subscriptions or access through their institution are able to view the full article.