Giulia is a postdoc in the Department of Computing working on Reinforcement Learning. Prior to that she was a PhD student in the Department of Computer Science at UCL, where she worked on the interplay between Optimal Transport and Machine Learning. She has focused on theoretical aspects of such interplay, when using Optimal Transport learning and averaging measures. Before the PhD, she did both BSc and MSc in pure mathematics at the University of Pavia in Italy and she spent two terms at Humboldt in Berlin.
Giulia's research interests are, broadly speaking, machine learning and optimal transport. She is now focusing on studying reinforcement learning algorithms that target medical treatment optimization. In particular, she's exploring two directions, that consist in studying measures of similarity between policies and in leveraging the flexibility of distributional reinforcement learning to design safe and efficient off-policy learning methods.