BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Causal Identification under Unmeasured Confounding: Opportunities 
 in Multidimensional Data - Linbo Wang (University of Toronto)
DTSTART:20260303T114500Z
DTEND:20260303T123000Z
UID:TALK244375@talks.cam.ac.uk
DESCRIPTION:Unmeasured confounding is typically viewed as a fundamental ob
 stacle to causal identification. At the same time\, modern datasets are in
 creasingly multidimensional\, featuring large collections of exposures and
  multiple outcomes measured jointly and over time. In this talk\, I argue 
 that such multidimensional structure can substantially reshape the identif
 ication problem. Structural sparsity among many exposures can help disting
 uish causal effects from spurious associations. Likewise\, symmetry and re
 dundancy across multiple or longitudinal outcomes can introduce constraint
 s that facilitate identification despite confounding.&nbsp\;From this pers
 pective\, identifiability is not only a property of isolated treatment-out
 come pairs\, but emerges from the broader structural geometry of the data 
 generating process. Multidimensional designs thus provide new avenues for 
 discovering causal structure and restoring identification in settings wher
 e classical single treatment\, single outcome paradigms&nbsp\;would fail.
LOCATION:Seminar Room 1\, Newton Institute
END:VEVENT
END:VCALENDAR