Sparse kernel regression modelling using combined locally regularized orthogonal least squares and D-optimality experimental design
Sparse kernel regression modelling using combined locally regularized orthogonal least squares and D-optimality experimental design
The paper proposes an efficient nonlinear identification algorithm by combining a locally regularized orthogonal least squares (LROLS) model selection with a D-optimality experimental design. The proposed algorithm aims to achieve maximized model robustness and sparsity via two effective and complementary approaches. The LROLS method alone is capable of producing a very parsimonious model with excellent generalization performance. The D-optimality design criterion further enhances the model efficiency and robustness. An added advantage is that the user only needs to specify a weighting for the D-optimality cost in the combined model selecting criterion and the entire model construction procedure becomes automatic. The value of this weighting does not influence the model selection procedure critically and it can be chosen with ease from a wide range of values.
1029-1036
Chen, S.
9310a111-f79a-48b8-98c7-383ca93cbb80
Hong, X.
b8f251c3-e142-4555-a54c-c504de966b03
Harris, C.J.
c4fd3763-7b3f-4db1-9ca3-5501080f797a
June 2003
Chen, S.
9310a111-f79a-48b8-98c7-383ca93cbb80
Hong, X.
b8f251c3-e142-4555-a54c-c504de966b03
Harris, C.J.
c4fd3763-7b3f-4db1-9ca3-5501080f797a
Chen, S., Hong, X. and Harris, C.J.
(2003)
Sparse kernel regression modelling using combined locally regularized orthogonal least squares and D-optimality experimental design.
IEEE Transactions on Automatic Control, 48 (6), .
Abstract
The paper proposes an efficient nonlinear identification algorithm by combining a locally regularized orthogonal least squares (LROLS) model selection with a D-optimality experimental design. The proposed algorithm aims to achieve maximized model robustness and sparsity via two effective and complementary approaches. The LROLS method alone is capable of producing a very parsimonious model with excellent generalization performance. The D-optimality design criterion further enhances the model efficiency and robustness. An added advantage is that the user only needs to specify a weighting for the D-optimality cost in the combined model selecting criterion and the entire model construction procedure becomes automatic. The value of this weighting does not influence the model selection procedure critically and it can be chosen with ease from a wide range of values.
Text
01205199.pdf
- Other
More information
Published date: June 2003
Additional Information:
submitted for publication in June 2002
Organisations:
Southampton Wireless Group
Identifiers
Local EPrints ID: 257755
URI: http://eprints.soton.ac.uk/id/eprint/257755
ISSN: 0018-9286
PURE UUID: d4f27fe7-87ab-4d3d-ab96-a9372e2868a0
Catalogue record
Date deposited: 26 Jun 2003
Last modified: 14 Mar 2024 06:03
Export record
Contributors
Author:
S. Chen
Author:
X. Hong
Author:
C.J. Harris
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