|
Jesper MØLLERModern spatial point process modelling and inferenceThis lecture will review modern spatial point process modelling and inference based on recent advances in computational statistics.Spatial point processes are used to model point patterns where the points typically are positions or centres of objects in a two- or three-dimensional region. The points may be decorated with marks (such as sizes or types of the objects) whereby marked point processes are obtained. The areas of applications are manifold and include astronomy, ecology, forestry, geography, image analysis, and spatial epidemiology. For more than 30 years spatial point processes have been a major area of research in spatial statistics. It is expected that research in spatial point processes will continue to be of importance as new technology makes huge amounts of spatial point process data available and new applications emerge. Moreover, in the last 15 years computational methods, and particularly Markov Chain Monte Carlo (MCMC) methods, have undergone major developments. In fact, some of the earliest applications of computational methods in statistics are related to spatial point processes. Some material to be present can be found in the monograph belowI ; other can be found in recent publications mentioned at my webpage;
Moller, J. and Waagepetersen, R. (2003). Statistical inference and
simulation for spatial point processes. Chapman and Hall/CRC Press.
|
|
|
|
|
|
|
|
|