Extreme density level set estimation for input--output functions via the adaptive splitting technique

Rudy Pastel (1, 2) , Jérôme Morio (3, 4) , François Le Gland (1)

1 ASPI - Applications of interacting particle systems to statistics
2 ONERA - The French Aerospace Lab [Châtillon]
3 GdR MASCOT-NUM - Méthodes d'Analyse Stochastique des Codes et Traitements Numériques
4 ONERA - The French Aerospace Lab [Toulouse]

In order to estimate the minimum volume set of a probability very close to zero, a plug-in estimator of density level set is built based on the adaptive splitting technique, a rare event dedicated probability technique. This algorithm is the main novelty of this paper. It is applied on the realistic estimation of a booster fallout zone.

Domaines: Probabilités [math.PR]

Rudy Pastel, Jérôme Morio, François Le Gland. Extreme density level set estimation for input--output functions via the adaptive splitting technique. Journal of computational science, 2015, 6, pp.40--46. ⟨10.1016/j.jocs.2014.11.001⟩. ⟨hal-01110391⟩