Title :

Asymptotically Exact Data Augmentation: Models, Properties, and Algorithms

Author(s) :

Vono, Maxime [Auteur]
Signal et Communications [IRIT-SC]
Dobigeon, Nicolas [Auteur]
Institut National Polytechnique (Toulouse) [Toulouse INP]
Institut Universitaire de France [IUF]
Signal et Communications [IRIT-SC]
Chainais, Pierre [Auteur]
Centre de Recherche en Informatique, Signal et Automatique de Lille - UMR 9189 [CRIStAL]

Journal title :

Journal of Computational and Graphical Statistics

Pages :

335-348

Publisher :

Taylor & Francis

Publication date :

2021

ISSN :

1061-8600

English keyword(s) :

Approximation
Auxiliary variables
Bayesian inference
Divide-and-conquer
Robustness

HAL domain(s) :

Informatique [cs]/Intelligence artificielle [cs.AI]
Informatique [cs]/Vision par ordinateur et reconnaissance de formes [cs.CV]
Physique [physics]/Physique [physics]/Analyse de données, Statistiques et Probabilités [physics.data-an]
Informatique [cs]/Apprentissage [cs.LG]
Informatique [cs]/Traitement des images [eess.IV]
Informatique [cs]/Traitement du signal et de l'image [eess.SP]

English abstract : [en]

Data augmentation, by the introduction of auxiliary variables, has become an ubiquitous technique to improve convergence properties, simplify the implementation or reduce the computational time of inference methods such ...
Show more >Data augmentation, by the introduction of auxiliary variables, has become an ubiquitous technique to improve convergence properties, simplify the implementation or reduce the computational time of inference methods such as Markov chain Monte Carlo ones. Nonetheless, introducing appropriate auxiliary variables while preserving the initial target probability distribution and offering a computationally efficient inference cannot be conducted in a systematic way. To deal with such issues, this article studies a unified framework, coined asymptotically exact data augmentation (AXDA), which encompasses both well-established and more recent approximate augmented models. In a broader perspective, this article shows that AXDA models can benefit from interesting statistical properties and yield efficient inference algorithms. In non-asymptotic settings, the quality of the proposed approximation is assessed with several theoretical results. The latter are illustrated on standard statistical problems.Show less >

Language :

Anglais

Peer reviewed article :

Oui

Audience :

Internationale

Popular science :

Non

ANR Project :

Artificial and Natural Intelligence Toulouse Institute

Collections :

• Centre de Recherche en Informatique, Signal et Automatique de Lille (CRIStAL) - UMR 9189

Source :

Harvested from HAL