GLMs and GAMs with Spatial, or Spatial- Temporal Correlation using inlabru

Onsite course :

GLMs and GAMs with Spatial, or Spatial-Temporal Correlation using INLA

 

This course offers a practical introduction to the analysis of spatial, and spatial-temporal data using generalised linear models (GLMs) and generalised additive models (GAMs). We begin with how to add spatial structure to regression models using frequentist techniques, and then introduce Bayesian methods, focusing on how to implement models with spatial and temporal dependency using INLA. The course covers a range of data types and distributions, including Gaussian, Poisson, generalised Poisson, negative binomial, Bernoulli, and Tweedie.

Participants will learn how to build models that incorporate spatial correlation, or spatio-temporal correlation, and how to address practical challenges such as modelling in the presence of natural barriers (e.g., coastlines, forests) that prevent spatial correlation from extending freely across space, such as marine/terrestrial boundaries or fragmented habitats. We will also cover the use of more complex spatial meshes and multivariate likelihoods to accommodate study areas with isolated groups of sites.

We will utilise the R package inlabru.

Through hands-on exercises, you will gain experience in fitting and interpreting models for continuous, count, and binary data, and understand how to adapt your models to the specific structure and distribution of your data. This is an applied and non-technical course that focuses on the practical implementation in R.

 

Course content

Module 1

  • General introduction.
  • Theory presentation on adding spatial dependence to a regression model using frequentist techniques.
  • One exercise showing how to add spatial dependency to a regression model using frequentist tools.
  • Brief introduction to Bayesian analysis. Conjugate priors. Diffuse versus informative priors.
  • Exercise showing how to execute a linear regression model in inlabru.

Module 2

  • Exercise showing how to add spatial correlation to a linear regression model using R-INLA and inlabru.
  • Exercise showing how to execute a Poisson GLM in inlabru.
  • Exercise showing how to add spatial correlation to a Poisson GLM.

Module 3

  • Short theory presentation on GAMs.
  • Exercise on executing a Gaussian GAM in R-INLA and and inlabru.
  • Exercise on adding spatial correlation to a Gaussian GLM/GAM
  • Time allowing:
  • Exercise showing how to add spatial correlation to a negative binomial GLM. With a barrier.
  • Exercise showing how to add spatial correlation to a Bernoulli GLM.

Module 4

  • Exercise showing how to add spatial-temporal correlation to a Poisson or negative binomial GLM/GAM.
  • Exercise showing how to add spatial-temporal correlation to a Tweedie GLM/GAM.
  • Time allowing:
  • Exercise showing how to add spatial-temporal correlation to a Bernoulli GLM/GAM.

We reserve the right to change the exercises. Pdf files of all theory material will be provided. All exercises consists of data sets and annotated R scripts. Access to the course website is for 12 months. The course website also contains on demand video.

 

Pre-required knowledge
Working knowledge of R, data exploration, linear regression and GLM (Poisson, negative binomial, Bernoulli). This is a non-technical course.

The course website provides preparatory materials, including on-demand videos and R scripts covering multiple linear regression, basic matrix notation, generalised linear models, model validation using DHARMa, and the explanation of variograms. If you are not familiar with these methods, please review them before the course begins.

General
Please ensure that you have system administration rights to install R, and R packages on your computer. Instructions what to install is on the course website.

1 hour face-to-face

  • The course includes a 1-hour face-to-face video chat with the instructors (to be used after the course). You are invited to apply the statistical techniques discussed during the course on your own data and if you encounter any problems, you can ask questions during the 1-hour face-to-face chat.

Discussion Board

  • A discussion board (access for 6 months) allows for interaction on course content between instructors and participants.
    In this course, we will only use the Gaussian, Poisson and negative binomial distributions.

For terms and conditions, see:
https://www.highstat.com/index.php/component/hikashop/checkout/termsandconditions/step-3/pos-6/tmpl-component