Show simple item record

dc.contributor.authorRubenthaler, Sylvain
HAL ID: 3728
dc.contributor.authorJacob, Pierre E.
HAL ID: 16651
ORCID: 0000-0002-4662-1051
dc.contributor.authorDoucet, Arnaud
dc.date.accessioned2013-10-23T08:13:49Z
dc.date.available2013-10-23T08:13:49Z
dc.date.issued2013
dc.identifier.urihttps://basepub.dauphine.fr/handle/123456789/11911
dc.language.isoenen
dc.subjectState-space modelsen
dc.subjectSmoothinen
dc.subjectSequential Monte Carloen
dc.subjectObserved information m atrixen
dc.subjectScore vectoren
dc.subjectMaximum likelihooden
dc.subject.ddc519en
dc.titleDerivative-Free Estimation of the Score Vector and Observed Information Matrix with Application to State-Space Modelsen
dc.typeDocument de travail / Working paper
dc.contributor.editoruniversityotherLaboratoire Jean Alexandre Dieudonné (JAD) http://math.unice.fr/ CNRS : UMR7351 – Université Nice Sophia Antipolis [UNS];France
dc.contributor.editoruniversityotherDepartment of Statistics http://www.stats.ox.ac.uk/ University of Oxford;Royaume-Uni
dc.description.abstractenIonides, King et al. (see e.g. Inference for nonlinear dynamical systems, PNAS 103) have recently introduced an original approach to perform maximum likelihood parameter estimation in state-space models which only requires being able to simulate the latent Markov model according its prior distribution. Their methodology relies on an approximation of the score vector for general statistical models based upon an artificial posterior distribution and bypasses the calculation of any derivative. Building upon this insightful work, we provide here a simple "derivative-free" estimator of the observed information matrix based upon this very artificial posterior distribution. However for state-space models where sequential Monte Carlo computation is required, these estimators have too high a variance and need to be modified. In this specific context, we derive new derivative-free estimators of the score vector and observed information matrix which are computed using sequential Monte Carlo approximations of smoothed additive functionals associated with a modified version of the original state-space model.en
dc.publisher.nameUniversité Paris-Dauphineen
dc.publisher.cityParisen
dc.identifier.citationpages21en
dc.identifier.urlsitehttp://fr.arxiv.org/abs/1304.5768en
dc.subject.ddclabelProbabilités et mathématiques appliquéesen
dc.description.submittednonen


Files in this item

FilesSizeFormatView

There are no files associated with this item.

This item appears in the following Collection(s)

Show simple item record