Inhomogeneous spatial point processes
June 2000
AbstractThe aim of this paper is to give a review of some recent
models derived
from homogeneous spatial point
processes with interaction. The focus is on
models derived from homogeneous
Markov point processes. For some of the
models, the interaction is
location dependent. A new type of approximate
transformation model with this
property is also suggested. The statistical
inference based on likelihood and
pseudolikelihood is discussed for the
different models. In particular,
it is shown that for transformation models,
the pseudolikelihood function can
be decomposed in a similar fashion as the
likelihood function.
download as : [ps-file] |