Poisson process models for a combination of points and counts in space

摘要: A spatial point process is a stochastic model determining the locations of events in some region C . Events may be nests breeding colony birds, tress forest, or cities country. One goal statistics to underlying and thus interpret complicated through parameter estimates based on known from processes. Techniques have been developed for estimating parameters pro­ cess, given data at either aggregate levels. However, it remains unclear how (i.e., counts sections) with subset exact events). This study investigates nonhomogeneous Poisson intensity function {A(s; 9) : 9 £ ©}. The de­ pend variable, location s alone, both. We propose mixture an accommodate both level information if possible. It turns out that proposed combined forms useful covariates are available. appears give better than does only count) data. shows more we know precise maximum likelihood become process. asymptotic properties estimator also studied.