Erwan Lecarpentier, a doctoral student at the Department of Complex System Engineering (DISC), has received the ISAE-SUPAERO Foundation thesis prize for his work on Reinforcement Learning in a Non-stationary Environment. A 2016 ISAE-SUPAERO graduate with ENSICA engineering training, Erwan prepared his doctoral thesis at ISAE-SUPAERO and ONERA under the supervision of Emmanuel Rachelson, Professor of Automatic Learning and Optimization at ISAE-SUPAERO, and Guillaume Infantes, researcher at JoliBrain (an artificial intelligence software and service company).
After defending his thesis in 2020, Erwan continued his work with Emmanuel Rachelson and is now looking for a postdoctoral position to pursue his career in academics. Reinforcement learning is a sub-branch of automatic learning that deals with solving the problem of sequential decision-making in discrete time. It is a question of automatically learning the optimum behavior for an agent interacting with an uncertain environment.
How do you teach machines to control complex problems? Erwan sums up his thesis for you in 3 minutes flat! This thesis prize was presented by the ISAE-SUPAERO Foundation for the excellence of the research carried out and the outlook for the thesis’ applications.