Improving Generalization in Meta Reinforcement Learning using Learned Objectives.

Louis Kirsch Sjoerd van Steenkiste Jürgen Schmidhuber

Published in: ICLR (2020)

Keyphrases

reinforcement learning
efficient learning
learned knowledge
machine learning
learning algorithm
artificial intelligence
reinforcement learning algorithms
state space
markov decision processes
function approximation
temporal difference
learning process
dynamic programming
least squares