Enviar aquest missatge de text: Robust reinforcement learning with Bayesian optimisation and quadrature