Safe Reinforcement Learning for Continuous Spaces through Lyapunov-Constrained Behavior.

Sigurd Aksnes Fjerdingen Erik Kyrkjebø

Published in: SCAI (2011)

Keyphrases

reinforcement learning
dynamical systems
action space
state space
function approximation
continuous state and action spaces
control theory
human behavior
neural network
optimal policy
adaptive control
reinforcement learning algorithms
real robot
sufficient conditions
multi agent
discrete data
learning algorithm