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


Current browse context:


Change to browse by:

References & Citations


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

Statistics > Machine Learning

Title: Rank-One NMF-Based Initialization for NMF and Relative Error Bounds under a Geometric Assumption

Abstract: We propose a geometric assumption on nonnegative data matrices such that under this assumption, we are able to provide upper bounds (both deterministic and probabilistic) on the relative error of nonnegative matrix factorization (NMF). The algorithm we propose first uses the geometric assumption to obtain an exact clustering of the columns of the data matrix; subsequently, it employs several rank-one NMFs to obtain the final decomposition. When applied to data matrices generated from our statistical model, we observe that our proposed algorithm produces factor matrices with comparable relative errors vis-\`a-vis classical NMF algorithms but with much faster speeds. On face image and hyperspectral imaging datasets, we demonstrate that our algorithm provides an excellent initialization for applying other NMF algorithms at a low computational cost. Finally, we show on face and text datasets that the combinations of our algorithm and several classical NMF algorithms outperform other algorithms in terms of clustering performance.
Comments: Accepted by the IEEE Transactions on Signal Processing
Subjects: Machine Learning (stat.ML); Information Theory (cs.IT); Methodology (stat.ME)
DOI: 10.1109/TSP.2017.2713761
Cite as: arXiv:1612.08549 [stat.ML]
  (or arXiv:1612.08549v2 [stat.ML] for this version)

Submission history

From: Zhaoqiang Liu [view email]
[v1] Tue, 27 Dec 2016 09:27:10 GMT (563kb,D)
[v2] Fri, 2 Jun 2017 08:04:07 GMT (3661kb,D)

Link back to: arXiv, form interface, contact.