Efficient Exploration in Continuous-time Model-based Reinforcement Learning

Reinforcement learning algorithms typically consider discrete-time dynamics, even though the underlying systems are often continuous in time.