UNIVERSITY OF GLASGOW

STATISTICS SEMINAR PROGRAMME

Wednesday, 26th April, 3pm

Inference for spatio-temporal processes

Noel CRESSIE (The Ohio State University)

Wednesday, 9th June, 3 pm

Optimal paired comparison designs in the case of
quadratic regression models

Emiel VAN BERKUM (TU Eindhoven, The Netherlands)

Wednesday, 16th June, 3 pm

The largest inclusions within a piece of steel

Clive W. ANDERSON (University of Sheffield)

Seminars take place in Room 1f(203), Mathematics Building,
University of Glasgow

For further information please contact the seminar organiser:

Ilya Molchanov
University of Glasgow : e-mail: [log in to unmask]
Department of Statistics : Ph.: + 44 141 330 5141
Glasgow G12 8QW : Fax: + 44 141 330 4814
Scotland, U.K. : http://www.stats.gla.ac.uk/~ilya/



   ABSTRACTS

INFERENCE FOR SPATIO-TEMPORAL PROCESSES

In this talk I propose models and inference for spatio-temporal
variability. The extension of traditional geostatistical methods,
such as kriging, to the spatio-temporal domain is one possible
approach to characterize this variability. One could describe this
approach as 'descriptive' and in this talk I contrast it with
a dynamic approach that exploits the unidirectional flow of time
in the presence of spatially colored noise. With the inclusion
of a measurement equation, this formulation leads naturally to
the development of a spatio-temporal Kalman filter. This can be
viewed as empirical Bayesian inference in a hierarchical model.
I then go on to discuss fully Bayesian inference for spatio-temporal
climatological problems.


OPTIMAL PAIRED COMPARISON DESIGNS IN THE CASE OF
QUADRATIC REGRESSION MODELS

In paired comparison experiments observations are made by comparing
objects. This implies that only the difference in response can be
observed. This method is used extensively in experimental situations
where objects can be judged only subjectively. In the case when a
regression model can be formulated, well known methods to construct
optimal designs can be extended to construct optimal paired comparison
designs.

In this talk the general principles of D-optimality of designs will be
given. These will be applied to construct D-optimal designs for paired
comparison experiments in the case of quadratic regression models.


THE LARGEST INCLUSIONS WITHIN A PIECE OF STEEL

Imagine a solid object, homogeneous except for the presence within
it of small particles of foreign material of different sizes.
Interest lies in the size of the largest of these particles, and how
that size relates to the volume of the solid. Direct observation
inside the solid is impossible, but particles intersecting the
surface can be seen in section.

This is a problem with particular relevance to new high quality
steels. All steels contain inclusions - small particles of impurity
- which influence fatigue strength. In the new so-called clean steels
the number and size of inclusions are much reduced and it becomes
important for quality-control and safety reasons to estimate the
likely size of the largest. Measurement of the cross-sections of
inclusions exposed in sampled polished surfaces of the steel can be
made reasonably routinely.

Without the emphasis on the largest particles, inference about
particle sizes on the basis of two-dimensional sections is a
standard problem in stereology (Wicksell's corpuscle problem of
1925). The talk will describe a development undertaken jointly with
Stuart Coles which concentrates specifically on inferences about
large particles, combining modern extreme value modelling with
stereological ideas. Both likelihood-based and Bayesian approaches
will be presented.

The problem highlights some general issues concerning the choice
of models in the analysis of extremes, which I also hope to explore.


%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%