Stanford Seminar - Towards Safe and Efficient Learning in the Physical World

April 5, 2024
Andreas Krause of ETH Zurich
How can we enable agents to efficiently and safely learn online, from interaction with the real world? I will first present safe Bayesian optimization, where we quantify uncertainty in the unknown objective and constraints, and, under some regularity conditions, can guarantee both safety and convergence to a natural notion of reachable optimum. I will then consider Bayesian model-based deep reinforcement learning, where we use the epistemic uncertainty in the world model to guide exploration while ensuring safety. Lastly I will discuss how we can meta-learn flexible probabilistic models from related tasks and simulations, and demonstrate our approaches on real-world applications, such as robotics tasks and tuning the SwissFEL Free Electron Laser.
About the speaker: inf.ethz.ch/people/person-det...
More about the course can be found here: stanfordasl.github.io/robotic...
View the entire AA289 Stanford Robotics and Autonomous Systems Seminar playlist: • Stanford AA289 - Robot...
► Check out the entire catalog of courses and programs available through Stanford Online: online.stanford.edu/explore