Safe Deep Reinforcement Learning by Verifying Task-Level Properties.

Enrico Marchesini Luca Marzari Alessandro Farinelli Christopher Amato

Published in: AAMAS (2023)

Keyphrases

reinforcement learning
levels of abstraction
desirable properties
learning algorithm
least squares
higher level
model checking
function approximation
temporal difference
real time
databases
low level
dynamic programming
reinforcement learning algorithms