|
Organizers |
Pivot Variables and Fiducial Inference: Application to Linear Mixed Models
by
Miguel Fonseca
Nova University of Lisbon
Coauthors: Joćo Tiago Mexia
Nova University of Lisbon
Uniquely determined fiducial densities defined as Radon-Nikodin derivatives are derived from pivot variables. The absolute maximum of such densities are introduced as plausible estimators and minimum measure confidence regions are obtained.
This approach is applied to normal mixed models. In the study of these models, variance components and estimable vectors are considered. A new optimal property is established for BLUE of estimable vectors.
Date received: January 30, 2008
Copyright © 2008 by the author(s). The author(s) of this document and the organizers of the conference have granted their consent to include this abstract in Atlas Conferences Inc. Document # cavi-66.