BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//wp-events-plugin.com//7.4.0.1//EN
TZID:America/New_York
X-WR-TIMEZONE:America/New_York
BEGIN:VEVENT
UID:842@fds.yale.edu
DTSTART;TZID=America/New_York:20250424T113000
DTEND;TZID=America/New_York:20250424T130000
DTSTAMP:20250916T142151Z
URL:https://fds.yale.edu/events/fds-colloquium-richard-samworth-cambridge-
 how-should-we-do-linear-regression/
SUMMARY:FDS Colloquium: Richard Samworth (Cambridge)\, "How should we do li
 near regression?"
DESCRIPTION:\nCohosted by Yale School of Public Health\n\n\n\nAbstract: In 
 the context of linear regression\, we construct a data-driven convex loss 
 function with respect to which empirical risk minimisation yields optimal 
 asymptotic variance in the downstream estimation of the regression coeffic
 ients. Our semiparametric approach targets the best decreasing approximati
 on of the derivative of the log-density of the noise distribution. At the 
 population level\, this fitting process is a nonparametric extension of sc
 ore matching\, corresponding to a log-concave projection of the noise dist
 ribution with respect to the Fisher divergence. The procedure is computati
 onally efficient\, and we prove that our procedure attains the minimal asy
 mptotic covariance among all convex M-estimators. As an example of a non-l
 og-concave setting\, for Cauchy errors\, the optimal convex loss function 
 is Huber-like\, and our procedure yields an asymptotic efficiency greater 
 than 0.87 relative to the oracle maximum likelihood estimator of the regre
 ssion coefficients that uses knowledge of this error distribution\; in thi
 s sense\, we obtain robustness without sacrificing much efficiency. \n\n\
 n\nSpeaker bio: Richard Samworth obtained his PhD in Statistics from the U
 niversity of Cambridge in 2004\, and has remained in Cambridge since\, bec
 oming a full professor in 2013 and the Professor of Statistical Science in
  2017. &nbsp\;His main research interests are in nonparametric and high-di
 mensional statistics\; he &nbsp\;has developed methods and theory for shap
 e-constrained inference\, missing data\, subgroup selection\, data perturb
 ation techniques (random projections\, subsampling\, the bootstrap\, knock
 offs)\, changepoint estimation and independence testing. Richard currently
  holds a European Research Council Advanced Grant. &nbsp\;He received the 
 COPSS Presidents' Award in 2018\, was elected a Fellow of the Royal Societ
 y in 2021 and served as co-editor of the Annals of Statistics (2019-2021).
 \n
CATEGORIES: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