SCOOL

The scientific project of Scool is focussed on sequential decision making under uncertainty. In particular, we study reinforcement learning and bandit problems. A particular point we would like to investigate is situations in which learning happens over long time spans; this typically involves changes in the environment (non stationary environments) in which the problem is sovled, changes in the problem itself, the function to optimize, the representation of the problem and the environment, ...

Our activities range from fundamental research to applications, and also collaborations and transfer with companies. We emphasize applications in sustainable development, or health.