BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Nonparametrics in causal inference: densities\, heterogeneity\, & 
 beyond - Edward Kennedy (Carnegie Mellon University)
DTSTART:20260127T133000Z
DTEND:20260127T143000Z
UID:TALK241516@talks.cam.ac.uk
DESCRIPTION:Much work in causal inference focuses on finite-dimensional ta
 rgets like average treatment effects. However\, many substantively importa
 nt causal questions involve inherently infinite-dimensional objects\, such
  as counterfactual outcome distributions\, heterogeneous treatment effect 
 surfaces\, and continuous treatment curves. These targets occupy a hybrid 
 space between classical parameter estimation and nonparametric function es
 timation. In this talk\, I survey some recent work involving these infinit
 e-dimensional causal estimands\, highlighting both model-based and model-f
 ree nonparametric approaches. I discuss how\, despite the impossibility of
  &radic\;n-rate estimation\, ideas from semiparametric theory (like double
  robustness) continue to play a central role. Throughout I emphasize the r
 elevance of these methods in social science applications.
LOCATION:Seminar Room 1\, Newton Institute
END:VEVENT
END:VCALENDAR