Title: Safe and Sound Reinforcement Learning
Abstract: Reinforcement Learning offers a rich paradigm for sequential decision-making. In the last few years, we have seen tremendous progress in terms of algorithms & applications. However most of these were restricted to low-risk settings, such as games and simulators. Yet the potential for real-world impact is tremendous, if we can ensure our models and policies are safe and sound. In this talk I will review some of the challenges we have seen with reproducibility, brittleness and instability of RL solutions. I will also discuss tools and methods recently proposed to avoid undesirable behaviors and increase our ability to tackle real-world problems, including in health-care settings.
Bio: Joelle Pineau is a faculty member at Mila and an Associate Professor and William Dawson Scholar at the School of Computer Science at McGill University, where she co-directs the Reasoning and Learning Lab. She is also co-Managing Director of Facebook AI Research, and the director of its lab in Montreal, Canada. She holds a BASc in Engineering from the University of Waterloo, and an MSc and PhD in Robotics from Carnegie Mellon University. Dr. Pineau’s research focuses on developing new models and algorithms for planning and learning in complex partially-observable domains. She also works on applying these algorithms to complex problems in robotics, health care, games and conversational agents. She serves on the editorial board of the Journal of Artificial Intelligence Research and the Journal of Machine Learning Research and is Past-President of the International Machine Learning Society. She is a recipient of NSERC’s E.W.R. Steacie Memorial Fellowship (2018), a Fellow of the Association for the Advancement of Artificial Intelligence (AAAI), a Senior Fellow of the Canadian Institute for Advanced Research (CIFAR), a member of the College of New Scholars, Artists and Scientists by the Royal Society of Canada, and a 2019 recipient of the Governor General’s Innovation Awards.