Dynamic Shielding for Reinforcement Learning in Black-Box Environments

no code implementations • 27 Jul 2022 • Masaki Waga, Ezequiel Castellano, Sasinee Pruekprasert, Stefan Klikovits, Toru Takisaka, Ichiro Hasuo

The dynamic shielding technique constructs an approximate system model in parallel with RL using a variant of the RPNI algorithm and suppresses undesired explorations due to the shield constructed from the learned model.

reinforcement-learning Reinforcement Learning (RL)