Studying for Doctor of Philosophy
Area of study:
New models for spatial point process data
Title:New models for spatial point process data
Supervisors: Dr Tilman Davies (Department of Statistics), Professor Martin Hazelton (Department of Statistics), Professor Adrian Baddeley (Curtin University)
- BA/BSc (University of Southern Queensland)
- BSc(Hons) Statistics (University of Southern Queensland)
A spatial point pattern refers to a dataset of spatial locations of events or things, which can occur in a diverse range of areas such as epidemiology, ecology and archaeology. Statistical analysis of such datasets tends to be concerned with identifying trends in the spatial arrangement of points. These processes are typically modelled using two components, one relating to deterministic trend, and one relating to stochastic dependence between points.
Distinguishing between these two components can be extremely important in order to properly understand the nature of the spatial variability, however, separating these components is very difficult due to unidentifiability. As a result, statistical methodology tends to focus on rigorously modelling either determinist trend or the stochastic dependence between points, not both simultaneously. This project aims to develop new methods in which deterministic and stochastic effects are combined but are identifiable. This would allow detailed inference to be performed on both components and result in a deeper understanding of the data.