BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Memento EPFL//
BEGIN:VEVENT
SUMMARY:Sparse and efficient methods for extreme events
DTSTART:20230209T143000
DTEND:20230209T153000
DTSTAMP:20260406T202302Z
UID:0471e43064c13014a0b58dbe0e57c897cdc62885c6a9206326ab950a
CATEGORIES:Conferences - Seminars
DESCRIPTION:Sebastian ENGELKE\, Université de Genève\nSeminar in Mathema
 tics\nAbstract: Natural hazards such as heat waves or flooding are driven 
 by very few rare events and may cause large costs in addition to their sev
 ere impacts on human health. The changing climate seems to alter the occur
 rence of such extreme scenarios even more strongly than the average values
  and an increasing number of record-shattering events occur around the wor
 ld. We present recent developments on how the risk of such extreme scenari
 os can be assessed more accurately in modern applications with high-dimens
 ional and complex data. The first class of methods is concerned with the o
 ut-of-sample prediction problem in machine learning\, where either the res
 ponse or the predictors in the test data go beyond the range of the traini
 ng data. The second class of methods relies on the notion of extremal grap
 hical models introduced in Engelke and Hitz (2020\, JRSSB). They enable th
 e analysis of complex rare events on network structures (e.g.\, floods) or
  large-scale spatial data (e.g.\, heat waves). We discuss data-driven stru
 cture learning algorithms that estimate graphs through L_1 penalization in
  possibly high dimensions. Recent results show that the underlying notion 
 of extremal conditional independence arises as a special case of a much mo
 re general theory for limits of sums and maxima of independent random vect
 ors.\n 
LOCATION:MA A1 10 https://plan.epfl.ch/?room==MA%20A1%2010 https://epfl.zo
 om.us/j/64026495609
STATUS:CONFIRMED
END:VEVENT
END:VCALENDAR