I build reproducible Bayesian models for spatial and spatio-temporal processes using INLA, inlabru, and MCMC to turn scattered observations into insights on how phenomena vary across space and time. For highly specialized and complex models, I custom-code Bayesian solvers using deterministic, stochastic, or hybrid approaches to explore posterior distributions.
My work focuses on computational statistics and Bayesian inference. These two pillars have allowed me to develop novel space-time models, focusing both on the classical analysis of species distribution models and on the spatio-temporal analysis of compositional data—particularly through my collaboration on the European project LAMASUS (Land Use and Management) for land-use analysis.
Fast, approximate Bayesian inference for latent Gaussian models.
Markov Chain Monte Carlo methods, diagnostics, and convergence checks.
SPDE-FEM approaches for building geostatistical, discrete graphical and point process models.
Hierarchical structures for irregular and complex real-world data.
Guides paired with working code and data, built so the material is directly reusable rather than purely descriptive.
Full education, research, and technical background.