BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Memento EPFL//
BEGIN:VEVENT
SUMMARY:Composite Mardinal Likelihood Estimation of Mixed Discrete Respons
 e Choice Models
DTSTART:20100414T111500
DTSTAMP:20260510T025827Z
UID:8bb01c4cd02f1d8b77c81348aaab20c11981c8d1d7be102f347a65f9
CATEGORIES:Conferences - Seminars
DESCRIPTION:Prof. Chandra Bhat\, Dept of Civil\, Architecture and Environm
 ental Engineering\, University of Texas at Austin\nThe likelihood function
 s of many discrete ordered and unordered-response choice models entail the
  evalua-tion of analytically-intractable integrals. For instance\, the use
  of a mixing mechanism to relax the independent and identically distribute
 d (IID) error term distribution in the multinomial logit model is well doc
 umented in the discrete choice literature on unordered multinomial respons
 e models. In such an approach\, the error term vector is effectively decom
 posed into an IID component vector and another vector of jointly distribut
 ed ran-dom coefficients that lends the non-IID structure. It is typical (t
 hough not always the case) to consider the joint distribution of the rando
 m coefficients to be normally distributed. A particular advantage of the m
 ixing approach is that it can be used for both cross-sectional choice data
  as well as panel data without any substan-tial conceptual and coding diff
 erence. However\, such mixed models also lead to intractable likelihood fu
 nc-tion expressions. Except in the case when the integration involves only
  1-2 dimensions\, maximum simulated likelihood (MSL) techniques are usuall
 y employed to estimate these models. Unfortunately\, for many practi-cal s
 ituations\, the computational cost to ensure good asymptotic MSL estimator
  properties can be prohibitive and literally infeasible as the number of d
 imensions of integration rises. Besides\, the accuracy of simulation techn
 iques is known to degrade rapidly at medium-to-high dimensions\, and the s
 imulation noise increases substantially. This leads to convergence problem
 s during estimation. In addition\, such simulation-based ap-proaches becom
 e impractical in terms of computation time\, or even infeasible\, as the n
 umber of mixing di-mensions grows. 
LOCATION:GC B3 424
STATUS:CONFIRMED
END:VEVENT
END:VCALENDAR