Minimizing the largest of the parameter variances. V([beta])-Optimality

  1. López Fidalgo, Jesús
Livre:
Model oriented data analysis: Proceedings of the 3rd International Workshop in Petrodvorets, Russia, May 25-30, 1992

Éditorial: Springer-Verlag

ISBN: 0-387-91457-9

Année de publication: 1993

Pages: 71-79

Type: Chapitre d'ouvrage

GOOGLE SCHOLAR

Résumé

We introduce a new criterion function that directly minimizes the variances of the estimators of the parameters. Covariances are ignored. Although the calculations are more laborious than in other more manageable criteria, such as that of D-optimality, we gain in fidelity to our objective.