BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Memento EPFL//
BEGIN:VEVENT
SUMMARY:Linear regression with unmatched data:    A deconvolution perspect
 ive
DTSTART:20230915T110000
DTEND:20230915T120000
DTSTAMP:20260407T103216Z
UID:75223a9aadb2d2dfa55c6c8ae601e2f5f34f6bdbf5232f7e1cea6487
CATEGORIES:Conferences - Seminars
DESCRIPTION:Prof. Fadoua Ballabdaoui\, ETH Zürich\, Switzerland\nAbstract
 :\nWe consider the regression problem where the response $Y\\in\\RR$ and t
 he covariate $X\\in\\RR^d$ for $d\\geq 1$ are unmatched. Under this scenar
 io\, one does not have access to pairs of observations from the distributi
 on of $(X\, Y)$\, but instead separate datasets $\\{Y_i\\}_{i=1}^n$ and $\
 \{X_j\\}_{j=1}^m$\, possibly collected from different sources. We study th
 is problem assuming that the regression function is linear and the noise d
 istribution is known or can be estimated. We introduce an estimator of the
  regression vector based on deconvolution and demonstrate its consistency 
 and asymptotic normality under an identifiability assumption. We show how 
 the method can be used in semi-supervised learning\, in case one has acces
 s to a small sample of matched pairs $(X_k\, Y_k)$. Some simulation result
 s as well as application to real datasets will be considered to illustrate
  the theory.\n\nBio:\nI obtained my PhD in Statistics in 2014 from the Dep
 artment of Statistics of University of Washington\, Seattle \nAfter a pos
 tdoc at the Institute of Mathematical Stochastics\, University of Goetting
 en\, I was appointed in 2016 as an Assistant Professor at Universite  Par
 is-Dauphine\, where I served until 2015.\nCurrently I am an Adjunct Profes
 sor at the Seminar for Statistics\, Department of Mathematics\, ETH Zurich
 . I am specialized in Non-parametric Statistics\, estimation under shape c
 onstraints and mixture models. I have grown a strong interest in AI techni
 ques\, with application in Art (especially poetry and music).\n 
LOCATION:ME C2 405
STATUS:CONFIRMED
END:VEVENT
END:VCALENDAR