Umberto Picchiniaffiliated with the university

Research areas and keywords


  • stochastic modelling, Bayesian inference, stochastic differential equations, state space models, hidden Markov models, biostatistics, Monte Carlo, computational statistics, mixed-effects hierarchical models


I am interested in inference for stochastic dynamical systems and computational statistics.

In particular, I am interested in statistically indentifying the parameters of complex, nonlinear, stochastic models, for example stochastic differential equations (SDEs) and state-space models.For this reason, I develop probabilistic Monte Carlo algorithms such as approximate Bayesian computation (ABC) methods, and more in general likelihood-free methods for models having "intractable likelihoods".

Applied work focuses on mathematical modelling of biomedical issues, such as mixed-effects models for pharmacokinetic/pharmacodynamic data, and other biophysical/biomedical problems, see below. 

I am the principal investigator for the interdisciplinary project entitled "Statistical Inference and Stochastic Modelling of Protein Folding" (here is an accessible description), funded by the Swedish Research Council (project id 2013-5167). The project is performed in collaboration with Professor Kresten Lindorff-Larsen (Dept. Biology, Copenhagen University), associate professor Julie Lyng Forman (Dept. Biostatistics, Copenhagen University) and PhD student Samuel Wiqvist (Centre for Mathematical Sciences, Lund University)

Recent research outputs

View All (19)