BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//132.216.98.100//NONSGML kigkonsult.se iCalcreator 2.20.4//
BEGIN:VEVENT
UID:20250624T140926EDT-9256aah5i5@132.216.98.100
DTSTAMP:20250624T180926Z
DESCRIPTION:Statistique\n\nTitle: Functional Peaks-over-threshold Analysis 
 and its Applications\n\nAbstract: Estimating the risk of single occurrence
 s of natural hazards has become impor- tant in recent decades\, but up unt
 il now it has been largely limited to re-using catalogues of historical ev
 ents\, which usually do not exceed 40 to 50 years in length\, and to numer
 ical models\, which require heavy computation and are often unreliable for
  extrapolation. Extreme value theory provides statistical methods for esti
 mating the frequency of past extreme events as well as for extrapolating b
 eyond observed severities\, but it has mostly been focused on studying uni
 vari- ate quantities. Consequently the majority of its applications to nat
 ural hazards have neglected their spatio-temporal characteristics. We pres
 ent an extension of peaks-over-threshold analysis to functions which allow
 s one to define complex extreme events as special types of exceedances\, a
 nd then obtain their limit tail distribution\, namely the generalized r-Pa
 reto process. We focus on a specific model based on log-Gaussian random fu
 nctions using classical covariance structures to characterize extremal dep
 endence. Then\, we describe a stochastic weather generator for extreme eve
 nts\, capable of quan- tifying the recurrence of past events as well as ge
 nerating completely new ones. The methodology is applied to several natura
 l hazards such as windstorms and rainfall.\n\nhttps://uqam.zoom.us/j/85224
 008878\n
DTSTART:20210218T200000Z
DTEND:20210218T210000Z
SUMMARY:Raphaël de Fondeville (Swiss Data Science Center\, EPFL)
URL:/mathstat/channels/event/raphael-de-fondeville-swi
 ss-data-science-center-epfl-328387
END:VEVENT
END:VCALENDAR