Andrea Cremaschi, Assistant Professor at the School of Science and Technology at IE University, was featured in the 2026 Biostatistical Seminar series of the Oslo Centre for Biostatistics and Epidemiology at the University of Oslo. His talk, “Latent Modularity in Multi-View Data,” addressed one of the central challenges in modern data analysis: how to cluster individuals when the available information comes from several heterogeneous sources.
In the seminar, Cremaschi presented a Bayesian framework for multi-view clustering in which each individual is assigned to a baseline cluster while each view can still depart from that baseline when the data call for a different grouping. This latent modularity structure offers a flexible way to preserve shared information across views without forcing all sources into the same partition.
The talk also examined the induced priors on view-specific cluster labels and partitions, clarifying the modelling consequences of the prior construction. The methodology is supported by a tailored Markov chain Monte Carlo algorithm and was illustrated through both simulation experiments and a clinical case study based on multi-view measurements from the GUSTO cohort.
This work was developed jointly with M. De Iorio, G. Page, and A. Jasra, and reflects Andrea Cremaschi’s ongoing research in Bayesian statistics, clustering, and graphical models.
Andrea Cremaschi is an assistant professor at IE SciTech School. He is also a member of IE Research Datalab, specializing in Bayesian Statistics, Clustering, and Graphical Models.