CÉROU, Frédéric; DEL MORAL, Pierre; FURON, Teddy; GUYADER, Arnaud

doi:10.1007/s11222-011-9231-6

Métadonnées

Licence d’utilisation du document

CÉROU, Frédéric
Applications of interacting particle systems to statistics [ASPI]
Institut de Recherche Mathématique de Rennes [IRMAR]

DEL MORAL, Pierre
Advanced Learning Evolutionary Algorithms [ALEA]

FURON, Teddy
Multimedia content-based indexing [TEXMEX]

Langue

Article de revue

Ce document a été publié dans

Statistics and Computing. 2012, vol. 22, n° 3, p. 795-908

Springer Verlag (Germany)

Résumé en anglais

This paper discusses a novel strategy for simulating rare events and an associated Monte Carlo estimation of tail probabilities. Our method uses a system of interacting particles and exploits a Feynman-Kac representation of that system to analyze their fluctuations. Our precise analysis of the variance of a standard multilevel splitting algorithm reveals an opportunity for improvement. This leads to a novel method that relies on adaptive levels and produces, in the limit of an idealized version of the algorithm, estimates with optimal variance. The motivation for this theoretical work comes from problems occurring in watermarking and fin- gerprinting of digital contents, which represents a new field of applications of rare event simulation techniques. Some numerical results show performance close to the idealized version of our technique for these practical applications.< Réduire

Mots clés en anglais

Rare event

Sequential importance sampling

Feynman-Kac formula

Metropolis-Hastings

Fingerprinting

Watermarking

DOI

http://dx.doi.org/10.1007/s11222-011-9231-6

Project ANR

Sécurité et fiabilité des techniques de tatouages - ANR-06-SETI-0009

Origine

Importé de hal

Unités de recherche

Institut de Mathématiques de Bordeaux (IMB) - UMR 5251