[Forum SIS] Seminario Chiara Tommasi

[Matteo Ruggiero]

=========================================================
		                 AVVISO DI SEMINARIO
=========================================================

Mercoledi' 24 Febbraio 2010, alle ore 14.30
presso il Dipartimento di Economia Politica e Metodi Quantitativi dell'Universita' di Pavia, 
via San Felice 5, aula del Consiglio di Facolta' (primo piano)


CHIARA TOMMASI
(Dipartimento di Scienze Economiche, Aziendali e Statistiche, Universita' degli Studi di Milano)


terra' un seminario dal titolo


"THE K-L OPTIMALITY CRITERION FOR DISCRIMINATING BETWEEN TWO RIVAL MODELS"


Abstract
One of the criticisms that are usually made of the general theory of optimal experimental design is that a particular model must be assumed before designing the experiment. Very often the experimenter has not just one but several possible models for the data to be collected. Then a model must be chosen after a discrimination hypothesis test.
The KL-optimality criterion, which is based on the popular Kullback-Leibler distance, enables us to choose the experimental conditions for discriminating between  two rival statistical models. When the rival models are nested, the Ds-criterion is another optimality criterion useful to this aim.
As an example, both Ds- and KL-optimum designs are computed for discriminating between univariate logistic regression models with or without random effects. Both these competing optimum designs are constructed numerically and in order to improve the  convergence speed of the numerical procedures some integral approximations are suggested.


Tutti gli interessati sono invitati a partecipare.

---------
Matteo Ruggiero
Department of Economics and Quantitative Methods
University of Pavia
Via San Felice 5, 27100, Pavia, Italy
Phone:+39.0382.98.6224
Fax:+39.0382.30.4226
Web:http://economia.unipv.it/ruggiero


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.stat.unipg.it/pipermail/sis/attachments/20100217/0cdff30b/attachment.html>