Elias Krainski from KAUST - Making complex interaction models easier to work with

par Elias Krainski (KAUST)

jeudi 25 juin 2026, 13:00 → 14:00 Europe/Paris

Module 1.4 (ISPED)

Speaker: Elias Krainski from KAUST
Title: Making complex interaction models easier to work with

Abstract: The Gaussian Markov Random Field (GMRF) model is a key ingredient to model dependency over different domains. It includes modeling temporal or spatial dependency and earlier statistical models leverage the GMRF used to model spatially varying trends. Then the most common approach became to consider a random interaction between two random effects, one to account for the temporal correlation and the other to account for the spatial dependence. This led to the use of the so-called Kronecker product models, when the precision matrix for the interaction random effect is the Kronecker between the one from the temporal domain and the one from the spatial domain. Our purpose is to make it easy to try different kinds of interaction models. By integrating recent advancements in both theoretical modeling and software architecture, our approach prioritizes model interpretability alongside computational accessibility. We leverage existing software ecosystems to provide an extensible suite of tools that streamlines the deployment of complex spatial and spatio-temporal models, including joint modeling for multivariate outcomes. The utility and performance of these tools are demonstrated through applications to real-world datasets. In particular, we show its use to jointly analyse the rates of multiple diseases.

 

Calendar subscription link for the complete seminar series:
https://indico.math.cnrs.fr/category/711/events.ics

Program of the Biostatistics seminars:
https://indico.math.cnrs.fr/category/711/

Subscribe to the seminar mailing list:
https://diff.u-bordeaux.fr/sympa/subscribe/seminaire.biostat.bph

Former e-seminars on our YouTube channel (mostly in French): https://www.youtube.com/channel/UCURp-hEQL7k23UzGfqgEurA/videos

 

Biostatistics seminar series from the Department of Public Health from the University of Bordeaux and the Bordeaux Population Health UMR 1219 research center