Note: David Balduzzi has transitioned from the institute (alumni).
I am interested in how neurons learn to cooperate effectively. To this end, I am developing geometric and information-theoretic techniques for analyzing neural networks and, more generally, distributed systems of classifiers.
