BEGIN:VCALENDAR
VERSION:2.0
CALSCALE:GREGORIAN
PRODID:iCalendar-Ruby
BEGIN:VEVENT
CATEGORIES:Class/ Workshop
DESCRIPTION:Jörg Stoye - Cornell University\n\nConfidence Interals for Proj
ections of Partially Identified Parameters (joint w/Hiroaki Kaido & Frances
ca Molinari)\n\nAbstract: We propose a bootstrap-based calibrated projecti
on procedure to build confidence intervals for single components and for sm
ooth functions of a partially identified parameter vector in moment (in)equ
ality models. The method controls asymptotic coverage uniformly over a larg
e class of data generating processes. The extreme points of the calibrated
projection confidence interval are obtained by extremizing the value of the
component (or function) of interest subject to a proper relaxation of stud
entized sample analogs of the moment (in)equality conditions. The degree of
relaxation\, or critical level\, is calibrated so that the component (or f
unction) of \, not itself\, is uniformly asymptotically covered with presp
ecified probability. This calibration is based on repeatedly checking feasi
bility of linear programming problems\, rendering it computationally attrac
tive. Nonetheless\, the program defining an extreme point of the confidence
interval is generally nonlinear and potentially intricate. We provide an a
lgorithm\, based on the response surface method for global optimization\, t
hat approximates the solution rapidly and accurately. The algorithm is of i
ndependent interest for inference on optimal values of stochastic nonlinear
programs. We establish its convergence under conditions satisfied by canon
ical examples in the moment (in)equalities literature. Our assumptions and
those used in the leading alternative approach (a profiling based method) a
re not nested. An extensive Monte Carlo analysis confirms the accuracy of t
he solution algorithm and the good statistical as well as computational per
formance of calibrated projection\, including in comparison to other method
s.
DTEND:20180918T171000Z
DTSTAMP:20180924T225506Z
DTSTART:20180918T154000Z
GEO:42.447319;-76.480995
LOCATION:Uris Hall\, 498
SEQUENCE:0
SUMMARY:Econometrics Workshop: Jörg Stoye
UID:tag:localist.com\,2008:EventInstance_3794175
URL:https://events.cornell.edu/event/econometrics_workshop_joerg_stoye
END:VEVENT
END:VCALENDAR