Being able to autonomously learn control with minimal prior knowledge is a key ability of intelligent systems. This has therefore always been a central focus in our research on neural reinforcement learning methods. A particular challenge in real world control scenarios are methods that are at the same time highly data-efficient and robust, since data-collection on real systems is time intensive and often expensive. I will discuss two main research areas that are crucial for progress towards this goal: highly efficient off-policy learning and effective exploration. I will give examples of learning agent designs that can learn increasingly complex tasks from scratch in simulation and reality.
Copyright. All Rights Reserved.