Improving Scalability of Reinforcement Learning by Separation of Concerns.

Harm van Seijen Mehdi Fatemi Joshua Romoff Romain Laroche

Published in: CoRR (2016)

Keyphrases

reinforcement learning
function approximation
fault tolerance
dynamic programming
state space
reinforcement learning algorithms
information systems
markov decision processes
neural network
search algorithm
optimal policy
transfer learning
learning tasks
partially observable
policy search
robotic control