BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Nonparametric maximum likelihood methods for binary response model
 s with random coefficients - Roger Koenker\, University College London
DTSTART:20190531T150000Z
DTEND:20190531T160000Z
UID:TALK119413@talks.cam.ac.uk
CONTACT:Dr Sergio Bacallado
DESCRIPTION:Single index linear models for binary response with random coe
 fficients have been extensively employed in many econometric settings unde
 r various parametric specifications of the distribution of the random coef
 ficients. Nonparametric maximum likelihood estimation (NPMLE) as proposed 
 by Cosslett (1983) and Ichimura and Thompson (1998)\, in contrast\, has re
 ceived less attention in applied work due primarily to computational diffi
 culties. We propose a new approach to computation of NPMLEs for binary res
 ponse models that significantly increase their computational tractability 
 thereby facilitating greater flexibility in applications. Our approach\, w
 hich relies on recent developments involving the geometry of hyperplane ar
 rangements\, is contrasted with the recently proposed deconvolution method
  of Gautier and Kitamura (2013). An application to modal choice for the jo
 urney to work in the Washington DC area illustrates the methods.\n\nJoint 
 work with Jiaying Gu
LOCATION:MR12
END:VEVENT
END:VCALENDAR