Robert McCarthy
5 Papers
Robert McCarthy is an academic researcher. The author has contributed to research in topics: Computer science. The author has an hindex of 1, co-authored 4 publications.
Chat about Author
Papers
Behaviour Discriminator: A Simple Data Filtering Method to Improve Offline Policy Learning
Qing Wang,Robert McCarthy,David Cordova Bulens,Kevin McGuinness,Noel E. O'Connor,Francisco Roldán Sánchez,Stephen J. Redmond +6 more
TL;DR: In this paper , the authors proposed a behaviour discriminator (BD) concept, a novel and simple data filtering approach based on semi-supervised learning, which can accurately discern expert data from a mixed-quality dataset.
Real Robot Challenge 2022: Learning Dexterous Manipulation from Offline Data in the Real World
Nicolas Gurtler,Felix Widmaier,Cansu Sancaktar,Sebastian Blaes,Pavel Kolev,Stefan Bauer,M. Wuthrich,Markus Wulfmeier,Martin Riedmiller,Arthur Allshire,Qiang Wang,Robert McCarthy,Han-Gyoo Kim,Jongchan Baek Pohang,Wookyong Kwon,Shanliang Qian,Yasunori Toshimitsu,M. Michelis,Amirhossein Kazemipour,Arman Raayatsanati,Hehui Zheng,Barnabas Gavin Cangan,Bernhard Schölkopf,Georg Martius +23 more
TL;DR: The rules of the competition are state, the methods used by the winning teams are presented and their results are compared with a benchmark of state-of-the-art offline RL algorithms on the challenge datasets.
Your Value Function is a Control Barrier Function: Verification of Learned Policies using Control Theory
TL;DR: In this paper , the authors propose to apply verification methods used in control theory to learned value functions and derive original theorems linking value functions to control barrier functions.
Winning Solution of Real Robot Challenge III
TL;DR: The winning solution of the real-robot phase of the Real Robot Challenge (RRC) 2022 as mentioned in this paper was a semi-supervised Behavioral Cloning (BC) algorithm that outperformed state-of-the-art RL algorithms.
Value Functions are Control Barrier Functions: Verification of Safe Policies using Control Theory
TL;DR: In this article , the authors propose a new approach to apply verification methods from control theory to learned value functions, which can be used to guarantee safe behavior of reinforcement learning policies, despite RL's generality and scalability.