BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//CERN//INDICO//EN
BEGIN:VEVENT
SUMMARY:Multivariate root-n-consistent smoothing parameter free matching e
 stimators and estimators of inverse density weighted expectations
DTSTART:20250605T090000Z
DTEND:20250605T101500Z
DTSTAMP:20260423T021500Z
UID:indico-event-12960@indico.math.cnrs.fr
DESCRIPTION:Speakers: Alexander MEISTER (Rostock University)\n\nExpected v
 alues weighted by the inverse of a multivariate density or\, equivalently\
 , Lebesgue integrals of regression functions with multivariate regressors 
 occur in various areas of applications\, including estimating average trea
 tment effects\, nonparametric estimators in random coefficient regression 
 models or deconvolution estimators in Berkson errors-in-variables models. 
 The frequently used nearest-neighbor and matching estimators suffer from b
 ias problems in multiple dimensions. By using polynomial least squares fit
 s on each cell of the Kth-order Voronoi tessellation for sufficiently larg
 e K\, we develop novel modifications of nearest-neighbor and matching esti
 mators which again converge at the parametric root-n-rate under mild smoot
 hness assumptions on the unknown regression function and without any smoot
 hness conditions on the unknown density of the covariates. We stress that 
 in contrast to competing methods for correcting for the bias of matching e
 stimators\, our estimators do not involve nonparametric function estimator
 s and in particular do not rely on sample-size dependent smoothing paramet
 ers. We complement the upper bounds with appropriate lower bounds derived 
 from information-theoretic arguments\, which show that some smoothness of 
 the regression function is indeed required to achieve the parametric rate.
  Simulations illustrate the practical feasibility of the proposed methods.
  This talk is based on a joint work with Hajo Holzmann (Philipps-Universit
 y of Marburg\, Germany).\nPaper\n\nhttps://indico.math.cnrs.fr/event/12960
 /
LOCATION:Auditorium 3 (Toulouse School of Economics )
URL:https://indico.math.cnrs.fr/event/12960/
END:VEVENT
END:VCALENDAR