BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Memento EPFL//
BEGIN:VEVENT
SUMMARY:Semiparametric Inference for Nonmonotone Missing-Not-at-Random Dat
 a: The No Self-Censoring Model
DTSTART:20210416T161500
DTEND:20210416T173000
DTSTAMP:20260408T041434Z
UID:7c291ca0563de357217903a1217c242045dfa0b9591c93eade698134
CATEGORIES:Conferences - Seminars
DESCRIPTION:Daniel Malinsky\, Mailman School of Public Health\, Columbia\n
 We study the identification and estimation of statistical functionals of m
 ultivariate data missing nonmonotonically and not-at-random\, taking a sem
 iparametric approach. Specifically\, we assume that the missingness mechan
 ism satisfies what has been previously called “no self-censoring” or 
 “itemwise conditionally independent nonresponse\,” which roughly corre
 sponds to the assumption that no partially observed variable directly dete
 rmines its own missingness status.\n\nWe show that this assumption\, combi
 ned with an odds ratio parameterization of the joint density\, enables ide
 ntification of functionals of interest\, and we establish the semiparametr
 ic efficiency bound for the nonparametric model satisfying this assumption
 . We propose a practical augmented inverse probability weighted estimator\
 , and in the setting with a (possibly high-dimensional) always-observed su
 bset of covariates\, our proposed estimator enjoys a certain double-robust
 ness property. We explore the performance of our estimator with simulation
  experiments and on a previously studied dataset of HIV-positive mothers i
 n Botswana.\n\nThis is joint work with Ilya Shpitser and Eric Tchetgen Tch
 etgen.\n 
LOCATION:https://epfl.zoom.us/j/82372051871
STATUS:CONFIRMED
END:VEVENT
END:VCALENDAR