Reinforcement learning (RL) is a branch of machine learning inspired by the way biological entities (e.g. humans, animals) learn. In recent years, we have seen how deep learning has revolutionized machine learning and reinforcement learning is no exception. By leveraging deep neural networks, RL agents have managed to achieve superhuman performance in some long-standing applications. In this talk, I will introduce deep RL and how I use it to respond to some environmental and societal challenges in my ongoing research at INRAE.

Voir le replay de la présentation