Jump to Navigation

15 décembre 2015

Indéfini
Heure et lieu: 
11h, salle de réunion, bâtiment 210
Nom intervenant: 
Edward IONIDES
Titre: 
Inference for dynamic and latent variable models via iterated, perturbed Bayes maps
Résumé: 

Iterated filtering algorithms are stochastic optimization procedures for latent variable models that recursively combine parameter perturbations with latent variable reconstruction. Previously, theoretical support for these algorithms has been based on the use of conditional moments of perturbed parameters to approximate derivatives of the log likelihood function. We introduce a new theoretical approach based on the convergence of an iterated Bayes map. A new algorithm supported by this theory displays substantial numerical improvement on the computational challenge of inferring parameters of a partially observed Markov process.

Année: 
2015
Organisme intervenant: 
University of Michigan, Department of Statistics
Date du jour: 
Mardi, Décembre 15, 2015


Main menu 2

Seminaire | by Dr. Radut