Sparse matrix decompositions for clustering
Sparse matrix decompositions for clustering
Clustering can be understood as a matrix decomposition problem, where a feature vector matrix is represented as a product of two matrices, a matrix of cluster centres and a matrix with sparse columns, where each column assigns individual features to one of the cluster centres. This matrix factorisation is the basis of classical clustering methods, such as those based on non-negative matrix factorisation but can also be derived for other methods, such as k-means clustering. In this paper we derive a new clustering method that combines some aspects of both, non-negative matrix factorisation and k-means clustering. We demonstrate empirically that the new approach outperforms other methods on a host of examples.
clustering, low-rank matrix approximation, sparsity, brain imaging
Blumensath, Thomas
470d9055-0373-457e-bf80-4389f8ec4ead
3 March 2014
Blumensath, Thomas
470d9055-0373-457e-bf80-4389f8ec4ead
Blumensath, Thomas
(2014)
Sparse matrix decompositions for clustering.
22nd European Signal Processing Conference (EUSIPCO'2014), Lisbon, Portugal.
01 - 05 Sep 2014.
5 pp
.
Record type:
Conference or Workshop Item
(Paper)
Abstract
Clustering can be understood as a matrix decomposition problem, where a feature vector matrix is represented as a product of two matrices, a matrix of cluster centres and a matrix with sparse columns, where each column assigns individual features to one of the cluster centres. This matrix factorisation is the basis of classical clustering methods, such as those based on non-negative matrix factorisation but can also be derived for other methods, such as k-means clustering. In this paper we derive a new clustering method that combines some aspects of both, non-negative matrix factorisation and k-means clustering. We demonstrate empirically that the new approach outperforms other methods on a host of examples.
Text
EUSIPCO2014.pdf
- Author's Original
More information
Published date: 3 March 2014
Venue - Dates:
22nd European Signal Processing Conference (EUSIPCO'2014), Lisbon, Portugal, 2014-09-01 - 2014-09-05
Keywords:
clustering, low-rank matrix approximation, sparsity, brain imaging
Organisations:
Signal Processing & Control Grp
Identifiers
Local EPrints ID: 364849
URI: http://eprints.soton.ac.uk/id/eprint/364849
PURE UUID: 9438c0c2-7a0c-4574-a3cf-044b4a75f8bd
Catalogue record
Date deposited: 13 May 2014 10:19
Last modified: 15 Mar 2024 03:34
Export record
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
