Training workshop on geospatial analytics using R and R-INLA November 27th- December 1st, 2023
Venue: AIMS Rwanda
About
This short course will provide a comprehensive introduction to concepts, methods, and R tools for geospatial data analytics, which involves collecting, exploring, modelling, and visualising data that exhibit dependencies in space and/or time. Specific focus will be given to Bayesian inference through the Integrated Nested Laplace Approximation (INLA) approach.
We will first go through the basics of Bayesian inference and will then learn how to model hierarchical structures. We will also introduce elements for geocomputation with R. Then we will move to the core of the course, by focusing on area level data and presenting how to model spatially structured random effects through conditional autoregressive specifications. Following that, we will extend the approach to include temporal dependency and touch briefly on spatio-temporal interactions. Moving on to geostatistical data we will introduce the stochastic partial differential equation (SPDE) approach, used for spatial modelling on a continuous field. We will then extend this to deal with spatio-temporal data. Finally, we will describe how to use R-INLA for more advanced problems in the spatio-temporal realm, for instance how to deal with misaligned data.
Throughout the course there will be practical examples from epidemiology, public health, environmental and social sciences fields. The course will be delivered through a combination of lectures and computer-based practical sessions.
Application deadline:
TBC
Final Selection:
TBC
Delivery dates:
November 27th- December 1st, 2023
Delivery format:
In Person and Online
Pre-requisites
It is recommended that people attending are familiar with R (https://www.r-project.org/). We will provide a list of R packages to install prior to the course.
Maximum number of participants: 35 students
We will inform you shortly about how to apply for this training.
Marta Blangiardo
I have a chair in Biostatistics and I am the Biostatistics and Data Science theme leader for the MRC Centre for Environment and Health. I have a PhD in applied statistics and work on different modelling aspects relevant to environmental exposures (air pollution, noise, climate) and health (mostly chronic but have recently started moving towards infectious diseases, particularly in low and middle income countries). I have been also working on ecology, specifically on mosquitos estimations and species distribution.
See my webpages https://www.imperial.ac.uk/people/m.blangiardo and https://marta.statistica.it/
Twitter: @martablangiardo
Monica Pirani
I am an assistant professor in Biostatistics and I am an investigator in the MRC Centre for Environment and Health. My research is drawn toward understanding the link between the changing human-environment relationship and its impact on people’s health. Methodologically, my research focuses on spatial and spatiotemporal statistics, time-series analysis, dimension reduction and methods for the integration of multi-scale sources of data.
I am also interested in Bayesian hierarchical modelling approaches, and in the exploitation of hierarchical modelling ideas into Bayesian nonparametrics.
See my webpage at https://www.imperial.ac.uk/people/monica.pirani
Twitter: @monpirani
Georges Bucyibaruta
I am a Research Associate in Biostatistics and Data Science for environmental epidemiology. My current research interests are based on developing and applying machine learning tools and Bayesian models useful for clustering environmental and community level variables to health outcomes; to understand and account for their distributional characteristics and their spatial, temporal, and spatial-temporal variation. I work on statistical methods to improve the characterisation of air pollution. I focus on several methodological aspects related to the link between multiple air pollution constituents and health.
See my webpage at https://www.imperial.ac.uk/people/g.bucyibaruta
Twitter: @GBucyibaruta
We acknowledge the generous financial support from our funder.
Inquiries and feedback:
AIMS NEI Research
research@nexteinstein.org
Monday (9.30-16.00) – November 27
Session 1.1: Introduction to Bayesian thinking (9.30-10.30)
Break (20min)
Session 1.2: Introduction to INLA and R-INLA (10.50-12.20)
Practical 1a (12.20-13.00)
Lunch (1 hr)
Session 1.3: Introduction to geospatial data (14.00-14.45)
Practical 1b (14.45-16.00)
Tuesday (9.30-15.30) – November 28
Session 1.3: Hierarchical models, prediction, prior specification (9.30-10.50)
Break (20min)
Practical 2a & Practical 2b (11.10-12.30)
Lunch (1 hr)
Session 2.1: Spatial models for small area data: disease mapping and ecological regression (13.30pm-14.30)
Practical 3b (14.30-15.30)
Wednesday (9.30-15.30) – November 29
Session 3.1: Introduction to temporal modelling (9.30-10.30)
Break (20min)
Practical 3a (10.50-11.15)
Session 3.2: Spatio-temporal models for small area data (11.15-12.30)
Lunch (1 hr)
Practical 3b (13.30-15:30)
Thursday (9.30-15.30) – November 30
Session 4.1: Introduction to Geostatistics (9.30-10.30)
Break (20min)
Session 4.2: Introduction to SPDE model with R-INLA (10.50-12.30)
Lunch (1 hr)
Practical 4a & Practical 4b (13.30-15.30)
Friday (9.30-15.30) – December 1
Session 5.1: Spatio-temporal model for geostatistical data (9.30-11.00)
Break (20 min)
Practical 5a (11.20-12.30)
Lunch (1 hr)
Afternoon session (13.30-15.30): Participants will have the possibility to give a short presentation about their work/research and get feedback from their peers and from the teaching team.
Organizing Institutions
- African Institute for Mathematical Sciences (AIMS), Research and Innovation Centre
- Imperial College of London
Scientific Organizing committee
- Dr. Paterne Gahungu
- Dr. Isambi Mbalawata
Local Organizing committee
- Dr. Isambi Mbalawata
- Mrs. Dative Tuyisenge
- Miss Alison Karungi
