Algorithms and literate programs for weighted low-rank approximation with missing data
Algorithms and literate programs for weighted low-rank approximation with missing data
Linear models identification from data with missing values is posed as a weighted low-rank approximation problem with weights related to the missing values equal to zero. Alternating projections and variable projections methods for solving the resulting problem are outlined and implemented in a literate programming style, using Matlab/Octave's scripting language. The methods are evaluated on synthetic data and real data from the MovieLens data sets.
978-3-642-16875-8
255-273
Markovsky, Ivan
7d632d37-2100-41be-a4ff-90b92752212c
January 2011
Markovsky, Ivan
7d632d37-2100-41be-a4ff-90b92752212c
Markovsky, Ivan
(2011)
Algorithms and literate programs for weighted low-rank approximation with missing data.
In,
Levesley, Jeremy, Iske, Armin and Georgoulis, Emmanuil
(eds.)
Approximation Algorithms for Complex Systems.
Springer, .
Record type:
Book Section
Abstract
Linear models identification from data with missing values is posed as a weighted low-rank approximation problem with weights related to the missing values equal to zero. Alternating projections and variable projections methods for solving the resulting problem are outlined and implemented in a literate programming style, using Matlab/Octave's scripting language. The methods are evaluated on synthetic data and real data from the MovieLens data sets.
Text
missing-data-2x1.pdf
- Accepted Manuscript
Archive
missing-data.tar
- Other
More information
Published date: January 2011
Additional Information:
Chapter: 12
Organisations:
Southampton Wireless Group
Identifiers
Local EPrints ID: 268296
URI: http://eprints.soton.ac.uk/id/eprint/268296
ISBN: 978-3-642-16875-8
PURE UUID: 34cf167b-c3c5-4670-8e86-c6a090a54e60
Catalogue record
Date deposited: 04 Dec 2009 14:48
Last modified: 14 Mar 2024 09:07
Export record
Contributors
Author:
Ivan Markovsky
Editor:
Jeremy Levesley
Editor:
Armin Iske
Editor:
Emmanuil Georgoulis
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