We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:


Current browse context:


Change to browse by:

References & Citations

DBLP - CS Bibliography


(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Computer Science > Information Theory

Title: Nearly Optimal Robust Subspace Tracking

Abstract: In this work, we study the robust subspace tracking (RST) problem and obtain one of the first two provable guarantees for it. The goal of RST is to track sequentially arriving data vectors that lie in a slowly changing low-dimensional subspace, while being robust to corruption by additive sparse outliers. It can also be interpreted as a dynamic (time-varying) extension of robust PCA (RPCA), with the minor difference that RST also requires a short tracking delay. We develop a recursive projected compressive sensing algorithm that we call Nearly Optimal RST via ReProCS (ReProCS-NORST) because its tracking delay is nearly optimal. We prove that NORST solves both the RST and the dynamic RPCA problems under weakened standard RPCA assumptions, two simple extra assumptions (slow subspace change and most outlier magnitudes lower bounded), and a few minor assumptions.
Our guarantee shows that NORST enjoys a near optimal tracking delay of $O(r \log n \log(1/\epsilon))$. Its required delay between subspace change times is the same, and its memory complexity is $n$ times this value. Thus both these are also nearly optimal. Here $n$ is the ambient space dimension, $r$ is the subspaces' dimension, and $\epsilon$ is the tracking accuracy. NORST also has the best outlier tolerance compared with all previous RPCA or RST methods, both theoretically and empirically (including for real videos), without requiring any model on how the outlier support is generated. This is possible because of the extra assumptions it uses.
Comments: A [short version](this http URL) will be presented at ICML 2018 (Long Talk). arXiv admin note: text overlap with arXiv:1803.00651
Subjects: Information Theory (cs.IT); Computer Vision and Pattern Recognition (cs.CV); Machine Learning (stat.ML)
Cite as: arXiv:1712.06061 [cs.IT]
  (or arXiv:1712.06061v4 [cs.IT] for this version)

Submission history

From: Praneeth Narayanamurthy [view email]
[v1] Sun, 17 Dec 2017 06:14:58 GMT (2397kb,D)
[v2] Mon, 12 Feb 2018 00:29:16 GMT (2404kb,D)
[v3] Sun, 25 Mar 2018 23:03:55 GMT (2755kb,D)
[v4] Fri, 6 Jul 2018 17:41:33 GMT (2763kb,D)

Link back to: arXiv, form interface, contact.