Imperial College Statistics Seminars
Friday, 29th January 1999. 2pm Room 140 Huxley Building,
180 Queen's Gate, South Kensington.
All welcome.
By: David Laws (University of Sheffield)
"Bayesian hierachical and nonparametric modelling for rare errors"
In this talk I will describe a Bayesian model for a scenario in which the
population of errors contains many
zeroes and there is a known covariate. This kind of structure typically occurs
in auditing, and I will use
auditing as the driving application of the method. The model is based on a
categorisation of the error
population together with a Bayesian non--parametric method of modelling errors
within some of the
categories. A hierarchical structure is introduced to model situations in which
the error distribution is
different under different levels of a factor, such as the different offices in
an organisation or company.
Inference is through simulation and examples are loosely based on data sets
provided by the National Audit
Office.
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|