Menu Content/Inhalt
Seminars Print
previous year previous month next month next year
See by year See by month See by week See Today Search Jump to month
Michel Broniatowski, Université Pierre et Marie Curie Print
Friday, 21 October 2011, 14:30 - 15:30

Michel Broniatowski, Université Pierre et Marie Curie

Weighted Sampling, Maximum Likelihood and Minimum Divergence Estimators

Abstract: This talk  explores Maximum Likelihood in parametric models in the context of Sanov type Large Deviation Probabilities. Connexion is stated with minimum divergence estimation under weighted sampling. It is shown that to any Cressie Read divergence it can be associated a specific weighted sampling scheme for which the resulting estimators are optimal in various respects. A link is established with Natural Exponential Families with power variance functions, unifying therefore various aspects of parametric inference. This paper makes an  use of variational forms of divergence criterions, as well as conditional large deviation results for empirical measures.

Location: Plaine
Contact: Jacqueline Bottemanne This e-mail address is being protected from spam bots, you need JavaScript enabled to view it

Back