School Seminars and Colloquia

Applied Asymptotics

Statistics Seminar

by Anthony Davison


Institution: Ecole Polytechnique Fédérale de Lausanne
Date: Thu 19th July 2007
Time: 12:00 PM
Location: Russell Love theatre, Richard Berry building, The University of Melbourne

Abstract: Most statistical modelling relies on asymptotic results to provide approximate confidence intervals,
significance probabilities, and the usual paraphenalia of inference. Usually the underlying basis is
the central limit theorem, but the resulting approximations can be poor or even misleading
in applications. The purpose of this talk is to show how much more accurate, even essentially
exact, approximations can be routinely found for many model types met in practice. The emphasis
will be on application rather than theory, and several examples will be given. The work, which
builds on contributions from many others, is joint with Alessandra Brazzale and Nancy Reid, and
the HOA package for the statistical environment R.

For More Information: Dr Owen Jones: odj@mailhost.ms.unimelb.edu.au