M. P. Frías Bustamante, A. Torres Signes, M. D. Ruiz Medina, J. Mateu Mahiques
A new approach is presented for spatial point pattern analysis in the Functional Data Analysis (FDA) framework. In this context, a new class of spatial Cox processes driven by a spatial Hilbert-valued random log-intensity is introduced. Its spatial correlation structure is estimated in a parametric framework in the spatial functional spectral domain. Specifically, an empirical statistical distance based on the spectral density and periodogram operators, inspired on Whittle functional estimation is considered. Strong-consistency of the parametric estimator is proved in the linear case. A simulation study is carried out to illustrate the performance of the proposed functional estimation methodology. Finally, we apply this methodology to the spatial functional prediction of respiratory disease mortality in the Spanish Iberian Peninsula, in the period 1980-2015.
Keywords: Infinite-dimensional log-intensity; Periodogram operator; Respiratory disease mortality; Spatial Autoregressive Hilbertian processes; Spatial Cox processes.
Scheduled
GT19 Time Space Statistics
June 7, 2022 12:00 PM
Conference hall
