Keynote Speakers
Prof. Diego Vidaurre
Centre de Recerca Matemàtica, Barcelona, Aarhus University, Oxford Universit
Keynote Topic:
"Machine learning and AI models as prototypes of perception"
Abstract:
A recurring but often glossed-over tension in the study of perception is that predicting a phenomenon and explaining it are not the same intellectual achievement — a distinction with a long history in the philosophy of science, but one that machine learning research frequently collapses. In this talk I revisit that distinction and ask what it means for computational models of perception. I will describe the two dominant machine-learning approaches to brain data: decoding and encoding. Both are intrinsically a predictive enterprise, exploiting whatever features of the signal have information. In this context, I will also present genephys, a generative model of the M/EEG signal that I developed, which sits between encoding and decoding. Finally, I will discuss how these methods can contribute to the actual understanding the mechanisms of vision, and how they can’t.
Biography:
Diego was a post-doctoral researcher at the Oxford Centre for Human Brain Activity (OHBA), where worked since 2012 to 2019. Since then, he holds an honorary position at the Department, moving as an Associate Professor to Aarhus University.He is interested in the development and application of computational methodology for characterising and understanding the complex network dynamics of activity in the brain, and for relating these to behaviour in a meaningful way. In particular, he aims to combine data from different sources (fMRI, MEG, EEG, anatomical data) using novel modelling techniques and state-of-the-art machine learning technology.
Prof. Guy Nagels
Professor of Digital Medicine, Vrije Universiteit Brussel
Keynote Topic:
"Conming soon..."
Biography:
As associate chair of neurology in the UZ Brussel, my clinical focus is diagnosis, treatment and rehabilitation in multiple sclerosis (MS). Cognition is an important though often underrated aspect of MS, and this has caused my research focus to grow into the modelling of cognition in MS and other neurodegenerative disorders. I feel that the combination of my clinical neuroscience background, together with my formal training as a computer science engineer, sets the stage for the development of more informed and more performant models of cognition. To this end, I have gathered a multidisciplinary team consisting of engineers, psychologists and fellow neurologists. Together we form the cognition and modelling group, part of the centre for neurosciences (C4N) of the Vrije Universiteit Brussel (VUB), where I am associate professor of cellular neurophysiology and coordinator of the neural engineering track in the master in biomedical engineering program. Our cognition and modelling group works in close collaboration with the Oxford centre for Human Brain Activity, where I hold academic visitor status since 2012, with the magnetoencephalography unit at the Université Libre de Bruxelles, and with industrial partners such as the spinoff neuroimaging company icometrix nv. We also have a long standing collaboration with the BIODEM lab at the University of Antwerp, resulting in publications on motor activity recordings and neurophysiological markers of cognition in dementia.
Keynote Topic:
""