Minimum-volume Rank-deficient Nonnegative Matrix Factorizations

Valentin Leplat, Andersen Man Shun Ang, Nicolas GillisICASSP 2019 · OpenReview

Abstract

In recent years, nonnegative matrix factorization (NMF) with volume regularization has been shown to be a powerful identifiable model; for example for hyperspectral unmixing, document classification, community detection and hidden Markov models. In this paper, we show that minimum-volume NMF (min-vol NMF) can also be used when the basis matrix is rank deficient, which is a reasonable scenario for some real-world NMF problems (e.g., for unmixing multispectral images). We propose an alternating fast projected gradient method for min-vol NMF and illustrate its use on rank-deficient NMF problems; namely a synthetic data set and a multispectral image.
BibTeX
@inproceedings{leplat2019minimum,
  author = {Valentin Leplat and Andersen Man Shun Ang and Nicolas Gillis},
  title = {{Minimum-volume Rank-deficient Nonnegative Matrix Factorizations}},
  booktitle = {IEEE International Conference on Acoustics, Speech and Signal Processing},
  year = {2019},
  url = {https://openreview.net/forum?id=FT2zp5zz5qB},
}

Loading related content...

Figure