Safety, Risk and Uncertainty in RL
Schedule
8:50 AM: Welcome and opening remarks
9:00 AM: Sergey Levine - Robust and Uncertainty-Aware Reinforcement Learning
9:45 AM: Contributed talk
Reinforcement Learning with Probabilistic Guarantees for Autonomous Driving
Maxime Bouton*, Jesper Karlsson, Alireza Nakhaei, Kikuo Fujimura, Mykel Kochenderfer and Jana Tumova
Risk-Sensitive Generative Adversarial Imitation Learning
Jonathan Lacotte, Mohammad Ghavamzadeh, Yinlam Chow* and Marco Pavone
Arushi Jain*, Ayush Jain and Doina Precup
10:30 AM: Coffee break & Posters
11:00 AM: Dorsa Sadigh - Learning Humans’ Preferences: A human-centered approach for safe interactions
11:45 AM: Contributed talk
Robust Bayes-Adaptive Planning under Model Uncertainty
Apoorva Sharma*, James Harrison and Marco Pavone
12:00 PM: Lunch break
1:30 PM: Angela Schoellig - Reinforcement Learning for Robotics: Provable Safety and Performance Guarantees by Combining Models and Data
2:15 PM: Contributed talk
Contextual Policy Optimisation
Supratik Paul*, Michael Osborne and Shimon Whiteson
2:30 PM: Jamie Morgenstern
3:15 PM: Contributed talk
A Fitted-Q Algorithm for Budgeted MDPs
Nicolas Carrara*, Romain Laroche, Tanguy Urvoy, Olivier Pietquin and Jean Leon Bouraoui
3:30 PM: Coffee break & Posters
4:00 PM: Philip Thomas - A New Type of Safe Reinforcement Learning
4:45 PM: Contributed talks
Safe Policy Improvement with Baseline Bootstrapping
Romain Laroche and Paul Trichelair*
Soft Safe Policy Improvement with Baseline Bootstrapping
Kimia Nadjahi, Romain Laroche, Rémi Tachet des Combes and Paul Trichelair*