Gaussian Processes and Reinforcement Learning for Identification and Control of an Autonomous Blimp | Synapse