Image

CSIRO’s Macroecological Modelling Research Group will be hosting Alain Zuur and Elena Leno who will present a five day course on GAM and GAMM using both Bayesian and frequentist approaches.

This course follows on from the very successful LME/GLMM course that was run by Alain and Elena last year.

It will focus on ecological problems/datasets and will be valuable to those who need to model non-linear patterns within their study species or system.
 
You can download the course flyer and register through Dr Zuur’s website http://www.highstat.com/statscourse.htm
               
Cost: Approx. $1140 (AUD).

Course content

Monday

• Revision multiple linear regression, Poisson GLM and Bernoulli GLM.
• Revision mixed effects models.
• Revision Bayesian statistics and MCMC.

Tuesday and Wednesday morning

• Theory presentation on GAM using the gam function from mgcv.
• Gaussian, Poisson, negative binomial and Bernoulli GAM using the mgcv package in R.
• Interactions between smoothers.
• This module is based on Chapter 2 in Zuur (2012).

Wednesday afternoon and Thursday morning

• Theory presentation on technical details of GAM
• GAMM for nested count and binary data (Poisson, negative binomial, Bernoulli) using gamm and gamm4.
• This module is based on various chapters in Zuur et al. (2014) and Chapter 3 in Zuur (2012).

Thursday afternoon and Friday morning

• Fitting a GAM or GAMM using MCMC in JAGS.

Friday afternoon

• What to write in a paper.
• Time allowing: Fitting a GAM (in INLA) and adding spatial and temporal correlation.
• A video file with fully worked out R code will be provided.