In this paper, lower bounds and upper bounds are given for the mass assigned to a set of maximal cliques in self-consistent estimates of CDF NPMLEs for multivariate (including univariate) interval censored data under the assumption that the censoring mechanism is ignorable for the purpose of likelihood inference. The bounds are applied to give upper bounds of the diameter and size of the polytope of CDF NPMLEs for multivariate censored data.
The paper is available in the following formats:
Research Unit on Children's psychosocial Maladjustment,
University of Montreal,
Montreal (Quebec) H3T 1J7
Alain C. Vandal
Faculty of Health and Environmental Sciences
Auckland University of Technology (AUT)
North Shore Campus
Private Bag 92006
|Alain C. Vandalfirstname.lastname@example.org|
Send any remarks to email@example.com.