DescriptionSpatial data, information collected from spatial locations, demands high-level statistical methodology to explore, investigate and make scientific conclusions. Such data arises in a wide range of applied fields such as aerial image processing, astronomy, ecology, engineering, environmental sciences, epidemiology, forestry, mineral prospecting, spatial economics and transportation. Spatial dtatistics, the statistical basis for spatial data science, encompasses statistical analysis of three different types of spatial data: geostatistical data, lattice data and point pattern data. Modeling of these different types of data requires different probabilistic and statistical tools.
This unit begins with a basic introduction to the three types of spatial data and develops some of the statistical tools required to describe and model such data. Then it moves on to in-depth study of a number of topics from the list: spatial stochastic processes, one and higher dimensional point processes, random fields, spatial covariance, variograms, stationarity and non-stationarity, kriging and spatial interpolation, first- and second-order intensity functions, summary functions, spatial models and estimation theory, simulation, spatial regression, spatio-temporal modeling, Bayesian methods in spatial statistics, and analysis of events on linear networks.
The unit will cover real world examples from many different fields. For statistical analysis and simulation the freeware package R will be used.