UNIVERSITY OF ST ANDREWS
Statistics Seminars - Candlemas Semester (March/April) 1999
____________________________________________________________
MONDAY 8 MARCH
Dr Chris JONES (Open University)
"Two separate topics: minimum distance estimation and local dependence"
MONDAY 15 MARCH
Professor Bill GURNEY (University of Strathclyde)
"Self-organisation, scale and stability in a spatial predator-prey interaction"
MONDAY 26 APRIL
Professor Frank BALL (University of Nottingham)
"MCMC for hidden continuous time Markov chains"
____________________________________________________________
All seminars start at 4 p.m., tea being available from 3.40 p.m. Visitors
will be very welcome.
All the seminars will be held in Lecture Theatre B, Mathematical Institute.
Further information from:
Dr I B J Goudie (email: [log in to unmask])
SEMINAR ABSTRACTS
Dr Chris JONES (The Open University)
"Two separate topics: minimum distance estimation and local dependence"
This seminar will comprise two short talks. The first will be about a new
class of minimum distance methods for robust estimation (Basu, Harris,
Hjort & Jones, Biometrika, 1998). The particular appeal of this class,
which forms a bridge from maximum likelihood to minimum L2 distance
estimation and beyond, is that although based on divergences involving
densities (rather than distribution functions), no member of the class
requires smoothing for its implementation. Some musings on the place of
this approach relative to others will be given.
The second mini-seminar will consider the concept of a local
dependence function. The usual measures of
dependence/correlation/association are scalars which, in some sense,
average out possibly different degrees of dependence in different regions
of sample space to give a single overall measure. By contrast, this
concept is a function which reflects different degrees of dependence in
different regions (Holland & Wang, Communications in Statistics, 1987;
Jones, Biometrika, 1986). Some properties of the function will be
described, and the potential for practical application indicated.
Professor Bill GURNEY (University of Strathclyde)
"Self-organisation, scale and stability in a spatial predator-prey interaction"
This talk will examine a number of elements in the behavioural repertoire
of a spatial extension of the Rosenzweig-McArthur model with immobile prey
and diffusively dispersing predators. The talk will start by discussing the
properties of invasions in one and two dimensional spaces. I shall place
particular emphasis on the role of regeneration after local extinction and
show how the prevention of unbiological local regrowth can be an important
determinant of system behaviour. After developing a series of analytic
approximations for the spatial and temporal scales of the invasion wave, I
shall illustrate how understanding invasions allows us to understand the
behaviour and scale of persistent patterns including spiral waves.
The first section of the talk will conclude by examining what the
theory of invasions and spiral waves can tell us about the mechanisms of
persistence in predator-prey interactions with highly oscillatory local
dynamics. This will lead to a discussion of an alternative route to
persistence, through the formation of self-organised heterogeneity in the
prey distribution. A series of numerical and analytic approximations will
be developed for the critical scales of such self-organised structures and
these will be contrasted with the comparable scales calculated for
persistent distributions produced by invasion-like dynamics.
Professor Frank BALL (University of Nottingham)*
"MCMC for hidden continuous time Markov chains"
Hidden Markov models have proved to be a very flexible class of models,
with many and diverse applications. Recently Markov chain Monte Carlo
(MCMC) techniques have provided powerful computational tools to make
inferences about the parameters of hidden Markov models, and about the
unobserved Markov chain, when the chain is in discrete time. In this talk,
I present a general algorithm, based on reversible jump MCMC for inference
in hidden Markov models where the unobserved chain runs in continuous time.
The method is illustrated with two examples. One is a relatively simple
application to Markov modulated Poisson processes. The second is a more
complex problem of inference from single ion channel data, and serves to
demonstrate the power and flexibility of the algorithm.
* Based on joint work with Tony O'Hagan, Yuzhi Cai and Jay Kadane.
--------------------------------------------------------------------------------
------------------------
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|