Jump to Navigation

25 janvier 2016

Indéfini
Heure et lieu: 
11h, salle de réunion, bâtiment 210
Nom intervenant: 
Sophie DONNET
Titre: 
Bayesian estimation for multidimensional Hawkes processes
Résumé: 

Multidimensional Hawkes processes are used to modelise multivariate neuron spike data. The estimation of intensity functions allows to understand the neuronal interaction structure. In a non-parametric frequentist framework, LASSO estimators have been proposed in the literature. In this work, we propose a Bayesian non-parametric estimation. We sample the posterior distribution through a Sequential Monte Carlo algorithm, well adapted to point processes.

Année: 
2016
Organisme intervenant: 
AgroParisTech-INRA, UMR 518 MIA
Date du jour: 
Lundi, Janvier 25, 2016


Main menu 2

Seminaire | by Dr. Radut