Backward elimination methods for associative memory network pruning
Backward elimination methods for associative memory network pruning
Three hybrid data based model construction/pruning formula are introduced by using backward elimination as automatic postprocessing approaches to improved model sparsity. Each of these approaches is based on a composite cost function between the model fit and one of three terms of A-/D-optimality / (parameter 1-norm in basis pursuit) that determines a pruning process. The A-/D-optimality based pruning formula contain some orthogonalisation between the pruned model and the deleted regressor. The basis pursuit cost function is derived as a simple formula without need for an orthogonalisation process. These different approaches to parsimonious data based modelling are applied to the same numerical examples in parallel to demonstrate their computational effectiveness.
90-99
Hong, X.
0a733642-067b-46e5-84db-f610140c22cb
Harris, C.J.
c4fd3763-7b3f-4db1-9ca3-5501080f797a
Brown, M.
28735aef-658a-4120-853f-29e986c94fc4
Chen, Sheng
9310a111-f79a-48b8-98c7-383ca93cbb80
13 September 2004
Hong, X.
0a733642-067b-46e5-84db-f610140c22cb
Harris, C.J.
c4fd3763-7b3f-4db1-9ca3-5501080f797a
Brown, M.
28735aef-658a-4120-853f-29e986c94fc4
Chen, Sheng
9310a111-f79a-48b8-98c7-383ca93cbb80
Hong, X., Harris, C.J., Brown, M. and Chen, Sheng
(2004)
Backward elimination methods for associative memory network pruning.
International Journal of Hybrid Intelligent Systems, 1 (2), .
(doi:10.3233/HIS-2004-11-211).
Abstract
Three hybrid data based model construction/pruning formula are introduced by using backward elimination as automatic postprocessing approaches to improved model sparsity. Each of these approaches is based on a composite cost function between the model fit and one of three terms of A-/D-optimality / (parameter 1-norm in basis pursuit) that determines a pruning process. The A-/D-optimality based pruning formula contain some orthogonalisation between the pruned model and the deleted regressor. The basis pursuit cost function is derived as a simple formula without need for an orthogonalisation process. These different approaches to parsimonious data based modelling are applied to the same numerical examples in parallel to demonstrate their computational effectiveness.
Text
backward
- Accepted Manuscript
Restricted to Repository staff only
Request a copy
Text
backward.pdf
- Other
More information
e-pub ahead of print date: February 2004
Published date: 13 September 2004
Organisations:
Southampton Wireless Group
Identifiers
Local EPrints ID: 258873
URI: http://eprints.soton.ac.uk/id/eprint/258873
PURE UUID: d1e97657-a13c-4611-a863-19c5bbb81b20
Catalogue record
Date deposited: 23 Feb 2004
Last modified: 14 Mar 2024 06:14
Export record
Altmetrics
Contributors
Author:
X. Hong
Author:
C.J. Harris
Author:
M. Brown
Author:
Sheng Chen
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