Analysis of massive marked point patterns with stochastic partial differential equations

摘要: Abstract In this paper we describe a novel approach to modelling marked point patterns based on recent computational developments for Bayesian inference. We use the flexible class of log-Gaussian Cox Processes model intensity different observed patterns. propose several types models account spatial variability and provide framework that allows common component all processes (regardless mark) also mark-specific components. way, method assessing whether share distribution or there are specific features. order fit these models, have resorted Integrated Nested Laplace Approximation (INLA) Stochastic Partial Differential Equation (SPDE) approach. This defines connection between process geostatistics, as pattern by means continuous process. Our new is applied massive dataset occurrence tornados in United States. divided 1950–2013 period according their magnitude fitted our proposed models.