BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//AMSE//Event Calendar//FR
CALSCALE:GREGORIAN
METHOD:PUBLISH
BEGIN:VEVENT
UID:event-8914@amse-aixmarseille.fr
DTSTAMP:20260430T142113Z
CREATED:20260430T142113Z
LAST-MODIFIED:20260430T142113Z
STATUS:CONFIRMED
SEQUENCE:0
SUMMARY:big data and econometrics seminar - Tim Verdonck
DTSTART:20220524T120000Z
DTEND:20220524T133000Z
DESCRIPTION:Predictive models are increasingly being used to optimize decis
 ion-making and minimize costs. A conventional approach is predict-then-opt
 imize: first\, a predictive model is built\; then\, this model is used to o
 ptimize decision-making. A drawback of this approach\, however\, is that it
  only incorporates costs in the second stage. Conversely\, the predict-and
 -optimize approach proposes learning a predictive model by directly minimi
 zing the cost of the downstream decision-making task. This is achieved by u
 sing a task-specific loss function incorporating the costs of different out
 comes in the first stage\, with the eventual aim of obtaining more cost-eff
 ective decisions in the second stage.  Fraud detection can be acknowledged
  as an (instance-dependent) cost-sensitive classification problem\, where t
 he costs due to misclassification vary between instances\, and requiring ad
 apted approaches for learning a classification model. In this presentation
 \, we present some classifiers that directly minimize an (instance-dependen
 t) cost measure when learning a classification model. The methods are evalu
 ated on real data.\\n\\nContact: Michel Lubrano: michel.lubrano[at]univ-amu
 .frPierre Michel: pierre.michel[at]univ-amu.fr\n\nPlus d'informations: htt
 ps://amse-aixmarseille.fr/en/events/tim-verdonck-4
LOCATION:Îlot Bernard du Bois - Salle 21\, AMU - AMSE\, 5-9 boulevard Maur
 ice Bourdet\, 13001 Marseille
URL;VALUE=URI:https://amse-aixmarseille.fr/en/events/tim-verdonck-4
CONTACT:Michel Lubrano: michel.lubrano[at]univ-amu.frPierre Michel:&nbsp\;p
 ierre.michel[at]univ-amu.fr
TRANSP:OPAQUE
END:VEVENT
END:VCALENDAR