Where and when

speaker: Silvia Cotroneo 
affiliation: Aalto University, Department of Neuroscience and Biomedical Engineering
date: Wednesday, May 27, 2026 – 11.00 am (Italian time) 
venue: Genoa (Italy), Polo Valletta Puggia – University of Genoa: DIBRIS-DIMA (room 322) 
link: Google Meet

Bits of information theory in MEEG signal analysis

abstract: After a brief review of differential entropy and mutual information, and of their role in independent component analysis (ICA), we will stretch the interpretation of mutual information as a similarity measure. This perspective will be examined in particular in the context of neural evoked responses, and compared with Pearson correlation. To conclude, an information-theoretic extension of canonical correlation analysis will be proposed, formulated as a multiset extension of ICA
through a mutual-information coupling term linking corresponding components across datasets.