Hypoelliptic diffusions: filtering and inference from complete and partial observations - Université Grenoble Alpes Accéder directement au contenu
Article Dans Une Revue Journal of the Royal Statistical Society: Series B Année : 2019

Hypoelliptic diffusions: filtering and inference from complete and partial observations

Résumé

The statistical problem of parameter estimation in partially observed hypoel-liptic diffusion processes is naturally occurring in many applications. However, due to the noise structure, where the noise components of the different coordinates of the multi-dimensional process operate on different time scales, standard inference tools are ill conditioned. In this paper, we propose to use a higher order scheme to discretize the process and approximate the likelihood, such that the different time scales are appropriately accounted for. We show consistency and asymptotic normality with non-typical convergence rates. When only partial observations are available, we embed the approximation into a filtering algorithm for the unobserved coordinates, and use this as a building block in a Stochastic Approximation Expectation Maximization algorithm. We illustrate on simulated data from three models; the Harmonic Oscillator, the FitzHugh-Nagumo model used to model the membrane potential evolution in neuroscience, and the Synaptic Inhibition and Excitation model used for determination of neuronal synaptic input.
Fichier principal
Vignette du fichier
hyposdesubmitted.pdf (525.76 Ko) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)
Loading...

Dates et versions

hal-01627616 , version 1 (02-11-2017)
hal-01627616 , version 3 (26-12-2018)
hal-01627616 , version 2 (12-08-2019)

Identifiants

Citer

Susanne Ditlevsen, Adeline Samson. Hypoelliptic diffusions: filtering and inference from complete and partial observations. Journal of the Royal Statistical Society: Series B, 2019, 81 (2), pp.361-384. ⟨10.1111/rssb.12307⟩. ⟨hal-01627616v3⟩
328 Consultations
340 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More