Importance sampling for online variational learning - Equipe Signal, Statistique et Apprentissage
Pré-Publication, Document De Travail Année : 2024

Importance sampling for online variational learning

Résumé

This article addresses online variational estimation in state-space models. We focus on learning the smoothing distribution, i.e. the joint distribution of the latent states given the observations, using a variational approach together with Monte Carlo importance sampling. We propose an efficient algorithm for computing the gradient of the evidence lower bound (ELBO) in the context of streaming data, where observations arrive sequentially. Our contributions include a computationally efficient online ELBO estimator, demonstrated performance in offline and true online settings, and adaptability for computing general expectations under joint smoothing distributions.
Fichier principal
Vignette du fichier
22-onlinevariational_hal.pdf (511.39 Ko) Télécharger le fichier
Origine Fichiers produits par l'(les) auteur(s)

Dates et versions

hal-04435065 , version 1 (02-02-2024)

Identifiants

Citer

Mathis Chagneux, Pierre Gloaguen, Sylvain Le Corff, Jimmy Olsson. Importance sampling for online variational learning. 2024. ⟨hal-04435065⟩
86 Consultations
54 Téléchargements

Altmetric

Partager

More