Inertial majorization-minimization algorithm for minimum-volume NMF
Inertial majorization-minimization algorithm for minimum-volume NMF
Nonnegative matrix factorization with the minimum-volume criterion (min-vol NMF) guarantees that, under some mild and realistic conditions, the factorization has an essentially unique solution. This result has been successfully leveraged in many applications, including topic modeling, hyperspectral image unmixing, and audio source separation. In this paper, we propose a fast algorithm to solve min-vol NMF which is based on a recently introduced block majorization-minimization framework with extrapolation steps. We illustrate the effectiveness of our new algorithm compared to the state of the art on several real hyperspectral images and document data sets.
Fast gradient method, Hyperspectral imaging, Majorization-minimization, Minimum volume, Nonnegative matrix factorization
1065-1069
European Signal Processing Conference, EUSIPCO
Thanh, Olivier Vu
e8eeb66d-10f7-4b01-bf11-e942b8b1ad18
Ang, Andersen
ed509ecd-39a3-4887-a709-339fdaded867
Gillis, Nicolas
76af3b6e-6ece-4191-a229-a7ff3616915f
Hien, Le Thi Khanh
68b76cac-dc01-4760-b909-998023a9f0b6
8 December 2021
Thanh, Olivier Vu
e8eeb66d-10f7-4b01-bf11-e942b8b1ad18
Ang, Andersen
ed509ecd-39a3-4887-a709-339fdaded867
Gillis, Nicolas
76af3b6e-6ece-4191-a229-a7ff3616915f
Hien, Le Thi Khanh
68b76cac-dc01-4760-b909-998023a9f0b6
Thanh, Olivier Vu, Ang, Andersen, Gillis, Nicolas and Hien, Le Thi Khanh
(2021)
Inertial majorization-minimization algorithm for minimum-volume NMF.
In 29th European Signal Processing Conference, EUSIPCO 2021 - Proceedings.
European Signal Processing Conference, EUSIPCO.
.
(doi:10.23919/EUSIPCO54536.2021.9616152).
Record type:
Conference or Workshop Item
(Paper)
Abstract
Nonnegative matrix factorization with the minimum-volume criterion (min-vol NMF) guarantees that, under some mild and realistic conditions, the factorization has an essentially unique solution. This result has been successfully leveraged in many applications, including topic modeling, hyperspectral image unmixing, and audio source separation. In this paper, we propose a fast algorithm to solve min-vol NMF which is based on a recently introduced block majorization-minimization framework with extrapolation steps. We illustrate the effectiveness of our new algorithm compared to the state of the art on several real hyperspectral images and document data sets.
This record has no associated files available for download.
More information
Published date: 8 December 2021
Venue - Dates:
29th European Signal Processing Conference, EUSIPCO 2021, , Dublin, Ireland, 2021-08-23 - 2021-08-27
Keywords:
Fast gradient method, Hyperspectral imaging, Majorization-minimization, Minimum volume, Nonnegative matrix factorization
Identifiers
Local EPrints ID: 495173
URI: http://eprints.soton.ac.uk/id/eprint/495173
ISSN: 2219-5491
PURE UUID: fbb6b683-187f-4636-afb3-473676e85d3c
Catalogue record
Date deposited: 31 Oct 2024 17:33
Last modified: 01 Nov 2024 03:05
Export record
Altmetrics
Contributors
Author:
Olivier Vu Thanh
Author:
Andersen Ang
Author:
Nicolas Gillis
Author:
Le Thi Khanh Hien
Download statistics
Downloads from ePrints over the past year. Other digital versions may also be available to download e.g. from the publisher's website.
View more statistics