Linear minimax estimation for random vectors with parametric uncertainty

Handle URI:
http://hdl.handle.net/10754/598720
Title:
Linear minimax estimation for random vectors with parametric uncertainty
Authors:
Bitar, E; Baeyens, E; Packard, A; Poolla, K
Abstract:
In this paper, we take a minimax approach to the problem of computing a worst-case linear mean squared error (MSE) estimate of X given Y , where X and Y are jointly distributed random vectors with parametric uncertainty in their distribution. We consider two uncertainty models, PA and PB. Model PA represents X and Y as jointly Gaussian whose covariance matrix Λ belongs to the convex hull of a set of m known covariance matrices. Model PB characterizes X and Y as jointly distributed according to a Gaussian mixture model with m known zero-mean components, but unknown component weights. We show: (a) the linear minimax estimator computed under model PA is identical to that computed under model PB when the vertices of the uncertain covariance set in PA are the same as the component covariances in model PB, and (b) the problem of computing the linear minimax estimator under either model reduces to a semidefinite program (SDP). We also consider the dynamic situation where x(t) and y(t) evolve according to a discrete-time LTI state space model driven by white noise, the statistics of which is modeled by PA and PB as before. We derive a recursive linear minimax filter for x(t) given y(t).
Citation:
Bitar E, Baeyens E, Packard A, Poolla K (2010) Linear minimax estimation for random vectors with parametric uncertainty. Proceedings of the 2010 American Control Conference. Available: http://dx.doi.org/10.1109/ACC.2010.5531063.
Publisher:
Institute of Electrical and Electronics Engineers (IEEE)
Journal:
Proceedings of the 2010 American Control Conference
KAUST Grant Number:
025478
Issue Date:
Jun-2010
DOI:
10.1109/ACC.2010.5531063
Type:
Conference Paper
Sponsors:
Supported in part by OOF991-KAUST US LIMITED underaward number 025478, the UC Discovery Grant ele07-10283 underthe IMPACT program, NASA Langley NRA NNH077ZEA001N,and NSF under Grant EECS-0925337.
Appears in Collections:
Publications Acknowledging KAUST Support

Full metadata record

DC FieldValue Language
dc.contributor.authorBitar, Een
dc.contributor.authorBaeyens, Een
dc.contributor.authorPackard, Aen
dc.contributor.authorPoolla, Ken
dc.date.accessioned2016-02-25T13:35:04Zen
dc.date.available2016-02-25T13:35:04Zen
dc.date.issued2010-06en
dc.identifier.citationBitar E, Baeyens E, Packard A, Poolla K (2010) Linear minimax estimation for random vectors with parametric uncertainty. Proceedings of the 2010 American Control Conference. Available: http://dx.doi.org/10.1109/ACC.2010.5531063.en
dc.identifier.doi10.1109/ACC.2010.5531063en
dc.identifier.urihttp://hdl.handle.net/10754/598720en
dc.description.abstractIn this paper, we take a minimax approach to the problem of computing a worst-case linear mean squared error (MSE) estimate of X given Y , where X and Y are jointly distributed random vectors with parametric uncertainty in their distribution. We consider two uncertainty models, PA and PB. Model PA represents X and Y as jointly Gaussian whose covariance matrix Λ belongs to the convex hull of a set of m known covariance matrices. Model PB characterizes X and Y as jointly distributed according to a Gaussian mixture model with m known zero-mean components, but unknown component weights. We show: (a) the linear minimax estimator computed under model PA is identical to that computed under model PB when the vertices of the uncertain covariance set in PA are the same as the component covariances in model PB, and (b) the problem of computing the linear minimax estimator under either model reduces to a semidefinite program (SDP). We also consider the dynamic situation where x(t) and y(t) evolve according to a discrete-time LTI state space model driven by white noise, the statistics of which is modeled by PA and PB as before. We derive a recursive linear minimax filter for x(t) given y(t).en
dc.description.sponsorshipSupported in part by OOF991-KAUST US LIMITED underaward number 025478, the UC Discovery Grant ele07-10283 underthe IMPACT program, NASA Langley NRA NNH077ZEA001N,and NSF under Grant EECS-0925337.en
dc.publisherInstitute of Electrical and Electronics Engineers (IEEE)en
dc.titleLinear minimax estimation for random vectors with parametric uncertaintyen
dc.typeConference Paperen
dc.identifier.journalProceedings of the 2010 American Control Conferenceen
dc.contributor.institutionMechanical Engineering, U.C. Berkeleyen
dc.contributor.institutionSystems Engineering and Automatic Control, Universidad de Valladoliden
kaust.grant.number025478en
All Items in KAUST are protected by copyright, with all rights reserved, unless otherwise indicated.