2003 Cornish Lecture

Date: Wednesday 15 October 2003

Speaker: Professor Adrian Baddeley, The University of Western Australia

Title: Practical analysis of spatial point patterns

Abstract:

The statistical analysis of spatial point patterns is an important and challenging task. It is important because of its many applications (e.g. in epidemiology, agriculture, ecology and materials engineering). It is challenging because it is methodologically different from most other areas of statistics. Statistical methods for point pattern data are relatively underdeveloped; the classical methods are very limited in scope.

In the last few years, it has finally become feasible to fit realistic, flexible models to real point pattern data. I will describe one new
model-fitting technique, and its implementation in R, and demonstrate it on real datasets.

Validation and criticism of a fitted model are still in their infancy. I will report some current research which defines residuals for spatial point processes, and diagnostic plots based on the residuals. These are also demonstrated on real data.

Biography:

Adrian Baddeley is Professor of Statistics at the University of Western Australia.

His research interests include stochastic geometry and its applications, spatial statistics, stereology, digital image analysis and statistical computing.

He has made outstanding contributions in these areas and is widely acknowledged as a leader in his field. He has received numerous honours including Australian Mathematical Society Medal in 1995 and the Hannan Medal for statistical Science in
2001.

He was the Bahadur Lecturer at the University of Chicago in 2002 and is a fellow of the Australian Academy of Science.