My research focuses on safe reinforcement learning, applied to behavioural planning for autonomous vehicles. Driving is a social game, as interactions between drivers result in complex coupled traffic dynamics. They are difficult to model manually and justify the need to learn from data. But to deploy a learning-based system in such a critical setting, a notion of risk must be estimated and controlled. However, this safety objective conflicts with that of efficiency, so both aspects must be adequately balanced.
PhD in Computer Science, expected 2020
Université de Lille
MSc in Executive Engineering, 2014