Between MDPs and semi-MDPs: A framework for temporal abstraction in reinforcement learning (1999)

by R S Sutton, D Precup, S P Singh
Venue:Artificial Intelligence