BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//wp-events-plugin.com//7.2.3.1//EN
TZID:America/New_York
X-WR-TIMEZONE:America/New_York
BEGIN:VEVENT
UID:875@fds.yale.edu
DTSTART;TZID=America/New_York:20250910T113000
DTEND;TZID=America/New_York:20250910T130000
DTSTAMP:20250916T142153Z
URL:https://fds.yale.edu/events/fds-colloquium-anay-mehrotra-yale/
SUMMARY:FDS Colloquium: Anay Mehrotra (Yale)\, "What Makes Treatment Effec
 ts Identifiable? Characterizations and Estimators Beyond Unconfoundedness"
DESCRIPTION:\nTalk summary: Sometimes\, like when studying the effects of s
 moking\, it is impossible to run a randomized control trial and we must re
 ly on observational data—where we observe who chose to receive treatment
  but do not control the assignment of the treatment. This leads to a funda
 mental challenge: we only see one outcome per individual (their health eit
 her as a smoker or non-smoker\, but never both)\, making it impossible to 
 directly observe causal effects. Unlike typical machine learning tasks whe
 re given enough data and computational power learning is always possible\,
  causal inference requires assumptions to enable learning. Unfortunately\,
  these assumptions frequently fail—severely limiting when causal effect
 s can be identified.\n\n\n\nIn this talk\, I will present a&nbsp\;characte
 rization of when treatment effects are identifiable from observational dat
 a\, which unifies existing approaches and enables exact identification of 
 treatment effects in&nbsp\;scenarios where current methods only provide ap
 proximate bounds. I will explain how these scenarios capture classical mod
 els such as special cases of sensitivity analysis and regression discontin
 uity designs. The characterization we present bridges causal inference wit
 h classical learning theory and opens exciting new avenues for causal esti
 mation when standard assumptions fail.\n\n\n\nThis talk is based on joint 
 work with Yang Cai\, Alkis Kalavasis\, Katerina Mamali\, and Manolis Zampe
 takis.\n\n\n\nSpeaker bio: Anay Mehrotra is a PhD candidate at Yale Univer
 sity advised by Amin Karbasi and Manolis Zampetakis. His research applies 
 learning-theoretic tools to missing-data challenges (ranging from causal i
 nference and truncated statistics to omissions driven by societal biases).
  He leverages these insights to assess the potential and limitations of mo
 dern AI systems\, often contributing novel tools back into learning theory
 . His work has received the Best Paper Award at COLT\, been featured in th
 e WIRED\, and received the Sir Binay Kumar Sinha award from IIT Kanpur. \
 n
CATEGORIES:Fellows Events,FDS Events,Colloquium
LOCATION:Yale Institute for Foundations of Data Science\, Kline Tower 13th 
 Floor\, Room 1327\, New Haven\, CT\, 06511\, United States
X-APPLE-STRUCTURED-LOCATION;VALUE=URI;X-ADDRESS=Kline Tower 13th Floor\, Ro
 om 1327\, New Haven\, CT\, 06511\, United States;X-APPLE-RADIUS=100;X-TITL
 E=Yale Institute for Foundations of Data Science:geo:0,0
END:VEVENT
BEGIN:VTIMEZONE
TZID:America/New_York
X-LIC-LOCATION:America/New_York
BEGIN:DAYLIGHT
DTSTART:20250309T030000
TZOFFSETFROM:-0500
TZOFFSETTO:-0400
TZNAME:EDT
END:DAYLIGHT
END:VTIMEZONE
END:VCALENDAR