Optimal monetary policy using reinforcement learning Discussion paper 51/2021: Natascha Hinterlang, Alina Tänzer