Design of Stochastic Machines Dedicated to Approximate Bayesian inferences

Abstract : We present an architecture and a compilation toolchain for stochastic machines dedicated to Bayesian inferences. These machines are not Von Neumann and code information with stochastic bitstreams instead of using floating point representations. They only rely on stochastic arithmetic and on Gibbs sampling to perform approximate inferences. They use banks of binary random generators which capture the prior knowledge on which the inference is built. The output of the machine is devised to continuously sample the joint probability distribution of interest. While the method is explained on a simple example, we show that our machine computes a good approximation of the solution to a problem intractable in exact inference.
Type de document :
Article dans une revue
IEEE Transactions on Emerging Topics in Computing, Institute of Electrical and Electronics Engineers, 2016, 〈10.1109/TETC.2016.2609926〉
Liste complète des métadonnées

Littérature citée [22 références]  Voir  Masquer  Télécharger

https://hal.archives-ouvertes.fr/hal-01374906
Contributeur : Pierre Bessiere <>
Soumis le : dimanche 2 octobre 2016 - 15:25:14
Dernière modification le : mardi 24 octobre 2017 - 10:20:01
Document(s) archivé(s) le : mardi 3 janvier 2017 - 13:34:20

Fichier

TETC-final.pdf
Fichiers produits par l'(les) auteur(s)

Identifiants

Citation

Marvin Faix, Raphä Laurent, Pierre Bessière, Emmanuel Mazer, Jacques Droulez. Design of Stochastic Machines Dedicated to Approximate Bayesian inferences. IEEE Transactions on Emerging Topics in Computing, Institute of Electrical and Electronics Engineers, 2016, 〈10.1109/TETC.2016.2609926〉. 〈hal-01374906〉

Partager

Métriques

Consultations de
la notice

351

Téléchargements du document

269