Approximate Bayesian computation with the sliced-Wasserstein distance - Archive ouverte HAL Accéder directement au contenu
Pré-Publication, Document De Travail Année : 2022

Approximate Bayesian computation with the sliced-Wasserstein distance

Résumé

Approximate Bayesian Computation (ABC) is a popular method for approximate inference in generative models with intractable but easy-to-sample likelihood. It constructs an approximate posterior distribution by finding parameters for which the simulated data are close to the observations in terms of summary statistics. These statistics are defined beforehand and might induce a loss of information, which has been shown to deteriorate the quality of the approximation. To overcome this problem, Wasserstein-ABC has been recently proposed, and compares the datasets via the Wasserstein distance between their empirical distributions, but does not scale well to the dimension or the number of samples. We propose a new ABC technique, called Sliced-Wasserstein ABC and based on the Sliced-Wasserstein distance, which has better computational and statistical properties. We derive two theoretical results showing the asymptotical consistency of our approach, and we illustrate its advantages on synthetic data and an image denoising task.
Fichier principal
Vignette du fichier
1910.12815.pdf (980.87 Ko) Télécharger le fichier
Origine : Fichiers éditeurs autorisés sur une archive ouverte

Dates et versions

hal-03945515 , version 1 (16-01-2024)

Identifiants

Citer

Kimia Nadjahi, Valentin de Bortoli, Alain Durmus, Roland Badeau, Umut Şimşekli. Approximate Bayesian computation with the sliced-Wasserstein distance. 2022. ⟨hal-03945515⟩
41 Consultations
17 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More