Learning state-action basis functions for hierarchical MDPs.

Sarah Osentoski Sridhar Mahadevan

Published in: ICML (2007)

Keyphrases

basis functions
reinforcement learning
state action
state space
learning algorithm
linear approximation
average reward
factored mdps
markov decision processes
linear combination
supervised learning
learning tasks
radial basis function
reward function
markov decision process
function approximators
semi supervised