J. Dvořák, M. Prokešová (2016)
Asymptotic properties of the minimum contrast estimators for projections of inhomogeneous space-time shot-noise Cox processes
Applications of Mathematics.
We consider a flexible class of space-time point process models—inhomogeneous shot-noise Cox point processes. They are suitable for modelling clustering phenomena, e.g. in epidemiology, seismology, etc. The particular structure of the model enables the use of projections to the spatial and temporal domain. They are used to formulate a stepwise estimation method to estimate different parts of the model separately. In the first step, the Poisson likelihood approach is used to estimate the inhomogeneity parameters. In the second and third steps, the minimum contrast estimation based on K-functions of the projected processes is used to estimate the interaction parameters. We study the asymptotic properties of the resulting estimators and formulate a set of conditions sufficient for establishing consistency and asymptotic normality of the estimators under the increasing domain asymptotics.ters of interest. Asymptotic normality of the two-step estimator with Palm likelihood is proved.
space-time point process, shot-noise Cox process, minimum contrast estimation, projection process, increasing domain asymptotics