[rl-list][workshops] The 1st International Workshop on Safe Reinforcement Learning Theory and its Applications @ IEEE MFI 2022
Josip Josifovski
Dear colleagues,
With apologies for cross-posting, we would like to invite you to the 1st international workshop on Safe Reinforcement Learning Theory and its Applications, co-located with the 2022 IEEE International Conference on Multisensor Fusion and Integration (MFI 2022). While the conference format is hybrid and the physical venue will be at Cranfield University in the United Kingdom, the workshop will be held entirely online and we welcome all researchers and students who are interested in safe RL to join us. Detailed workshop program and registration is available at https://saferl.online
Time and registration
The workshop will take place on 21st of September 2022 from 13:50 to 19:30 London time (GMT+1). To register for the workshop please fill in the registration form provided at https://saferl.online latest by 19th of September 2022. The workshop is open to the general public, exact details on how to join will be sent to your registration email after the registration is closed.
Motivation and Topics of Interest
Developing reinforcement learning (RL) algorithms that satisfy safety constraints is becoming increasingly important in real-world applications. How to ensure safety during RL applications is a challenging problem, which has received substantial attention in recent years. In this workshop we want to bring researchers from academia and industry and discuss how to address open issues in safe RL. The topics of interest include, but are not limited to:
Safe reinforcement learning (including single agent and multi-agent RL)
Metrics for safe real-world deployments
Robust reinforcement learning
Decision making under uncertainty
Safe robot learning dataset, simulators, and benchmarks
Robust perception, planning and control
Uncertainty quantification
Safety analysis and safety verification
Transfer learning
Sim2real transfer
Embodied systems and safety-critical applications
Explainability, transparency, and interpretability of learning
Repeatability and reliability of learning-based control
Safe multi-task learning and meta learning
Confirmed speakers
Sergey Levine (Associate Professor, University of California, Berkeley) - Safety in Reinforcement Learning by Leveraging Offline Data
Ding Zhao (Assistant Professor, Carnegie Mellon University) - Trustworthy Reinforcement Learning
Michael Everett (Assistant Professor, Northeastern University) - Certifiable Learning Machines
Simon Shaolei Du (Assistant Professor, University of Washington) - When are Offline Two-Player Zero-Sum Markov Games Solvable?
Martim Brandao (Assistant Professor, King's College London) - Are robots safe for everyone? The intersection of safety and fairness in robot motion
Chi Jin (Assistant Professor, Princeton University) - When Is Partially Observable Reinforcement Learning Not Scary?
Yali Du (Assistant Professor, King's College London) - Plug-and-Play Safe Multi-Agent Reinforcement Learning Using Constraint Augmentation
Zhehua Zhou (Postdoctoral Researcher, University of Alberta) - Safe Reinforcement Learning with Model Order Reduction Techniques
Ilias Kazantzidis (Ph.D Candidate, University of Southampton) - Human-in-the-loop Safe Reinforcement Learning
Workshop Advisors
Alois Knoll (Professor, Technical University of Munich)
Jun Wang (Professor, University College London)
Haitham Ammar (RL Research Group Leader, Huawei London)
Lei Ma (Associate Professor and Canada CIFAR AI Chair, University of Alberta)
Workshop Organizers
Shangding Gu (Ph.D. Candidate, Technical University of Munich)
Ding Zhao (Assistant Professor, Carnegie Mellon University)
Hao Dong (Assistant Professor, Peking University)
Long Yang (Postdoctoral Researcher, Peking University)
Meixin Zhu (Assistant Professor, Hong Kong University of Science and Technology (Guangzhou))
Josip Josifovski (Ph.D. Candidate, Technical University of Munich)
Mohammadhossein Malmir (Ph.D. Candidate, Technical University of Munich)
Guang Chen (Professor, Tongji University)
On behalf of the workshop organizers,
Josip Josifovski
Technical University of Munich (TUM)
Department of Informatics
Chair of Robotics, Artificial Intelligence and Real-time Systems (I6)
Schleißheimer Str. 90A, 85748
Garching bei München