Compressive System Identification in the Linear Time-Invariant framework

Handle URI:
http://hdl.handle.net/10754/597820
Title:
Compressive System Identification in the Linear Time-Invariant framework
Authors:
Toth, Roland; Sanandaji, Borhan M.; Poolla, Kameshwar; Vincent, Tyrone L.
Abstract:
Selection of an efficient model parametrization (model order, delay, etc.) has crucial importance in parametric system identification. It navigates a trade-off between representation capabilities of the model (structural bias) and effects of over-parametrization (variance increase of the estimates). There exists many approaches to this widely studied problem in terms of statistical regularization methods and information criteria. In this paper, an alternative ℓ 1 regularization scheme is proposed for estimation of sparse linear-regression models based on recent results in compressive sensing. It is shown that the proposed scheme provides consistent estimation of sparse models in terms of the so-called oracle property, it is computationally attractive for large-scale over-parameterized models and it is applicable in case of small data sets, i.e., underdetermined estimation problems. The performance of the approach w.r.t. other regularization schemes is demonstrated in an extensive Monte Carlo study. © 2011 IEEE.
Citation:
Toth R, Sanandaji BM, Poolla K, Vincent TL (2011) Compressive System Identification in the Linear Time-Invariant framework. IEEE Conference on Decision and Control and European Control Conference. Available: http://dx.doi.org/10.1109/CDC.2011.6160383.
Publisher:
Institute of Electrical and Electronics Engineers (IEEE)
Journal:
IEEE Conference on Decision and Control and European Control Conference
KAUST Grant Number:
025478
Issue Date:
Dec-2011
DOI:
10.1109/CDC.2011.6160383
Type:
Conference Paper
Sponsors:
Supported by NWO (grant no. 680-50-0927).Supported by NSF (grant no. ECCS-0925337) and OOF991-KAUST US LIMITED (award no. 025478).Supported by NSF (grant no. CNS-0931748).
Appears in Collections:
Publications Acknowledging KAUST Support

Full metadata record

DC FieldValue Language
dc.contributor.authorToth, Rolanden
dc.contributor.authorSanandaji, Borhan M.en
dc.contributor.authorPoolla, Kameshwaren
dc.contributor.authorVincent, Tyrone L.en
dc.date.accessioned2016-02-25T12:57:17Zen
dc.date.available2016-02-25T12:57:17Zen
dc.date.issued2011-12en
dc.identifier.citationToth R, Sanandaji BM, Poolla K, Vincent TL (2011) Compressive System Identification in the Linear Time-Invariant framework. IEEE Conference on Decision and Control and European Control Conference. Available: http://dx.doi.org/10.1109/CDC.2011.6160383.en
dc.identifier.doi10.1109/CDC.2011.6160383en
dc.identifier.urihttp://hdl.handle.net/10754/597820en
dc.description.abstractSelection of an efficient model parametrization (model order, delay, etc.) has crucial importance in parametric system identification. It navigates a trade-off between representation capabilities of the model (structural bias) and effects of over-parametrization (variance increase of the estimates). There exists many approaches to this widely studied problem in terms of statistical regularization methods and information criteria. In this paper, an alternative ℓ 1 regularization scheme is proposed for estimation of sparse linear-regression models based on recent results in compressive sensing. It is shown that the proposed scheme provides consistent estimation of sparse models in terms of the so-called oracle property, it is computationally attractive for large-scale over-parameterized models and it is applicable in case of small data sets, i.e., underdetermined estimation problems. The performance of the approach w.r.t. other regularization schemes is demonstrated in an extensive Monte Carlo study. © 2011 IEEE.en
dc.description.sponsorshipSupported by NWO (grant no. 680-50-0927).Supported by NSF (grant no. ECCS-0925337) and OOF991-KAUST US LIMITED (award no. 025478).Supported by NSF (grant no. CNS-0931748).en
dc.publisherInstitute of Electrical and Electronics Engineers (IEEE)en
dc.subjectCompressive Sensingen
dc.subjectLinear Time-Invariant Systemsen
dc.subjectSystem Identificationen
dc.titleCompressive System Identification in the Linear Time-Invariant frameworken
dc.typeConference Paperen
dc.identifier.journalIEEE Conference on Decision and Control and European Control Conferenceen
dc.contributor.institutionDelft University of Technology, Delft, Netherlandsen
dc.contributor.institutionColorado School of Mines, Golden, United Statesen
dc.contributor.institutionUC Berkeley, Berkeley, United Statesen
kaust.grant.number025478en
All Items in KAUST are protected by copyright, with all rights reserved, unless otherwise indicated.