BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Causal invariance and beyond - Ernst C. Wit (Università della Svi
 zzera italiana)
DTSTART:20260305T111500Z
DTEND:20260305T120000Z
UID:TALK244414@talks.cam.ac.uk
DESCRIPTION:As R.A. Fisher understood\, causal discovery is feasible when 
 controlled experiments on the system of interest can be conducted. Even in
  the absence of controlled human interventions\, reality often provides na
 tural experiments\, offering observations of the same system under slightl
 y shifted conditions. Rather than merely pooling such information\, we pro
 pose an approach that minimizes future prediction error by optimizing a fu
 nctional that accounts not only for pooled risk but also for risk differen
 tials across settings. We demonstrate how this method extends beyond the l
 inear regression framework into functional data analysis (FDA). This is jo
 int work with Lucas Kania (Carnegie Mellon)\, Philip Kennerberg\, Melania 
 Lembo (USI) and Veronica Vinciotti (Trento).
LOCATION:Seminar Room 1\, Newton Institute
END:VEVENT
END:VCALENDAR