Spatial Data Analysis

This course introduces statistical concepts and basic tools used in modelling and analysing spatial data: data on variables that are correlated in ‘space/location’ (geo-tagged data). Spatial data are commonly used in regional science and urban economics (related to property prices, crime, household income, etc.), epidemiology and public health (such as disease clusters, etc.), environmental science (air pollution, ozone density, etc.), ecology, biology, geology and other disciplines.

We start by looking at the practical aspects of organising and visualising spatial data where we will discuss methods available to organise and visualise (i) vector data and (ii) raster data using the R programming language. These are the two main ways in which spatial data are maintained. We also discuss how to use a coordinate reference system to give spatial awareness to a dataset and make meaningful maps including animated and interactive maps using OpenStreetMap and others.

We then consider the statistical aspects behind spatial data analysis with a special focus on how to re-align classical statistical methods towards spatial data. An outline is given to the three broad types of spatial data in spatial statistical analysis: (i) geostatistical data, (ii) areal data and (iii) point patterns. Standard spatial regression techniques are used to build models to explain attributes that are spatially correlated such as the number of COVID cases. These techniques are fully implemented using R throughout the course. You may follow the links to get a flavour of the applications in this course. SMU-X We will partner with a government statutory board to conduct a live project. Students, under the direct supervision of the faculty, will conduct a literature review, formulate research hypotheses, and conduct spatial data analyses to fulfil the main questions posed by the partner. The primary dataset will be provided by the partner. Depending on the formulated hypotheses, students may also need to collect additional data from publicly available sources. As the data provided by the partner will be confidential in nature, students are required to fulfil a confidentiality undertaking before embarking on the project. We will aim to complete the proposal by week 7. You are expected to present a progress report in week 11 and a final presentation in week 13.