Speaker
Finn Lindgren, Univ. of Edinburgh
Abstract
The INLA and inlabru R packages implement approximate Bayesian inferencefor latent Gaussian models, where spatial and spatio-temporal model components can be constructed from stochastic PDE models. I will discuss some of the practical considerations for the numerical methods, as well as explain some of the challenges and open problems for expanding these methods to larger and more general models.
Finn Lindgren: Embedding numerical stochastic PDE models in Bayesian inference for latent Gaussian models
Date: 2025-06-02
Time: 15:30 - 16:00