Mathematical and statistical modelling of complex systems is one of the key methods to understand and master their complexity. In this talk, I will first show how Bayesian machine learning can be combined with mathematical logic (specifically, temporal and spatio-temporal logic) to provide effective tools to analyse the behaviour of complex systems, unveiling their emergent properties, and to tackle system design, parameter synthesis, and model identification.
In the second part of the talk, I will discuss some work in progress, extending the previous ideas to tackle (related) problems like: automatic learning of model abstractions, using also Deep Neural Networks, with applications in multiscale modelling in systems biology and in geophysics, and learning of logic-based classifiers in the context of explainable AI.
Short bio: Luca Bortolussi is an Associate Professor of Computer Science at the Department of Mathematics and Geosciences of the University of Trieste, Italy, and guest professor of modelling and simulation at the department of Computer Science of the University of Saarland in Saarbruecken, Germany, where he worked from June 2014 to May 2015. Before that, Luca was assistant professor (Ricercatore) of Computer Science at the Department of Mathematics and Geosciences of the University of Trieste, Italy. Luca graduated in Mathematics at the University of Trieste in 2003 and got a PhD in Computer Science form the University of Udine in 2007. He has been an honorary fellow of the School of Informatics of the University of Edinburgh from 2013 to 2016, where he spent a sabbatical year in 2012. From 2012 to 2017 was an associate researcher at ISTI-CNR in Pisa.