CURSO: Using INLA to Fit Complex Spatial Models to Ecological Data
Charlotte Jones-Todd - Centre for Research into Ecological and Environmental Modelling - University of St. Andrews
- FCUL (DEIO) - Campo Grande - Bloco C6 Piso 4 - Sala: 6.4.31- 14:00h - 17:30h
- Quarta-feira, 1 de Abril de 2015
by Charlotte Jones-Todd
Centre for Research into Ecological and Environmental Modelling - University of St. Andrews
Many data sets are complex, resulting in complex models, e.g. complex spatial models. Usually Markov chain Monte Carlo (MCMC) methods have been used to fit such models. However, realistically complex models result in very long run times or are often impossible/unrealistic to fit. An alternative to MCMC is the Integrated Nested Laplace Approximation (INLA) approach which is much, much faster. The R-INLA software allows non-experts to fit complex models, and is suitable for a specific (but very large!) class of models. Coupled with the stochastic partial differential equation (SPDE) approach the methods presented here facilitate a fast and flexible way of constructing complex spatial and spatio-temporal models.
This short course provides a brief introduction to R-INLA and gives some examples modelling using the SPDE approach, through demonstrating how to fit complex spatial models to geo-statistical, lattice and point pattern data. While the theory behind INLA is complex and certainly interesting, the pitch of the course is applied, aiming at potential practitioners, focusing on the intuitive nature of the concepts and implementation, not on their theoretical properties.
14:00 – 14:15 Introduction & contents of “workshop”
14.15 – 15:00 INLA for ecological data
15:00 – 16:00 Demonstration & Practical session (INLA intro)
16:00 – 16:15 Coffee break
16:15 – 16:45 Brief introduction to the SPDE approach
16:45 – 17:15 SPDE practical session
17:15 – 17:30 Summary & close
Registration: mandatory, via e-mail email@example.com
Participants: Limited to 20 participants. Participants are expected to bring their own laptop with R and RINLA installed. Software and code will be provided to registered participants before the event.
Location: FCUL, DEIO, Building C6, room 6.4.31
Cost & Payment: no payment required, but a small contribution towards the coffee break snack will be most welcomme. A container for anonymous contributions for that purpose will be available on premisses.
Registration and further information:
Startfactor, Lda url: www.startfactor.pt
Curso em parceria com a "Startfactor - Statistical Consulting and Training"