Study of the Effective Number of Parameters in Nonlinear Identification Benchmarks

Anna Marconato; Maarten Schoukens; Yves Rolain; Joannes Schoukens

Study of the Effective Number of Parameters in Nonlinear Identification Benchmarks

Anna Marconato, Maarten Schoukens, Yves Rolain, Joannes Schoukens

Electricity

Research output: Chapter in Book/Report/Conference proceeding › Conference paper

14 Citations (Scopus)

Abstract

This paper discusses the importance of the notion of effective number of parameters as a measure of model complexity. Exploiting this concept allows a fair comparison of models obtained from different model classes. Several illustrative examples of linear and nonlinear models are presented to provide more insight in the problem. As one possible way of showing that model complexity can be reduced without having to pull any parameters to zero, an approach for rank reduced estimation based on the truncated SVD is also discussed. These ideas are then applied to two nonlinear real world problems: the Wiener-Hammerstein and the Silverbox benchmarks.

Original language	English
Title of host publication	52nd IEEE Conference on Decision and Control, Florence, Italy, December 10-13, 2013
Pages	4308-4313
Number of pages	6
Publication status	Published - 10 Dec 2013
Event	52nd IEEE Conference on Decision and Control - Firenze, Italy Duration: 10 Dec 2013 → 13 Dec 2013

Conference

Conference	52nd IEEE Conference on Decision and Control
Country/Territory	Italy
City	Firenze
Period	10/12/13 → 13/12/13

Keywords

nonlinear identification
bencjhmarks

SRP19: Strategic Research Programme: Center for model-based system improvement - From Computer-Aided Engineering to Model-Aided Engineering
Vandersteen, G., Rolain, Y., Wambacq, P., Kuijk, M., Vandersteen, G., Rolain, Y., Wambacq, P. & Kuijk, M.
1/11/12 → 31/10/24
Project: Fundamental

Cite this

@inproceedings{cfbb72d0b463497f8de677861a6a6765,

title = "Study of the Effective Number of Parameters in Nonlinear Identification Benchmarks",

abstract = "This paper discusses the importance of the notion of effective number of parameters as a measure of model complexity. Exploiting this concept allows a fair comparison of models obtained from different model classes. Several illustrative examples of linear and nonlinear models are presented to provide more insight in the problem. As one possible way of showing that model complexity can be reduced without having to pull any parameters to zero, an approach for rank reduced estimation based on the truncated SVD is also discussed. These ideas are then applied to two nonlinear real world problems: the Wiener-Hammerstein and the Silverbox benchmarks.",

keywords = "nonlinear identification, bencjhmarks",

author = "Anna Marconato and Maarten Schoukens and Yves Rolain and Joannes Schoukens",

year = "2013",

month = dec,

day = "10",

language = "English",

isbn = "978-1-4673-5716-6",

pages = "4308--4313",

booktitle = "52nd IEEE Conference on Decision and Control, Florence, Italy, December 10-13, 2013",

note = "52nd IEEE Conference on Decision and Control ; Conference date: 10-12-2013 Through 13-12-2013",

}

TY - GEN

T1 - Study of the Effective Number of Parameters in Nonlinear Identification Benchmarks

AU - Marconato, Anna

AU - Schoukens, Maarten

AU - Rolain, Yves

AU - Schoukens, Joannes

PY - 2013/12/10

Y1 - 2013/12/10

N2 - This paper discusses the importance of the notion of effective number of parameters as a measure of model complexity. Exploiting this concept allows a fair comparison of models obtained from different model classes. Several illustrative examples of linear and nonlinear models are presented to provide more insight in the problem. As one possible way of showing that model complexity can be reduced without having to pull any parameters to zero, an approach for rank reduced estimation based on the truncated SVD is also discussed. These ideas are then applied to two nonlinear real world problems: the Wiener-Hammerstein and the Silverbox benchmarks.

AB - This paper discusses the importance of the notion of effective number of parameters as a measure of model complexity. Exploiting this concept allows a fair comparison of models obtained from different model classes. Several illustrative examples of linear and nonlinear models are presented to provide more insight in the problem. As one possible way of showing that model complexity can be reduced without having to pull any parameters to zero, an approach for rank reduced estimation based on the truncated SVD is also discussed. These ideas are then applied to two nonlinear real world problems: the Wiener-Hammerstein and the Silverbox benchmarks.

KW - nonlinear identification

KW - bencjhmarks

M3 - Conference paper

SN - 978-1-4673-5716-6

SP - 4308

EP - 4313

BT - 52nd IEEE Conference on Decision and Control, Florence, Italy, December 10-13, 2013

T2 - 52nd IEEE Conference on Decision and Control

Y2 - 10 December 2013 through 13 December 2013

ER -

Study of the Effective Number of Parameters in Nonlinear Identification Benchmarks

Abstract

Conference

Keywords

Fingerprint

Projects

SRP19: Strategic Research Programme: Center for model-based system improvement - From Computer-Aided Engineering to Model-Aided Engineering

Cite this