Hierarchically Optimal Average Reward Reinforcement Learning.

Mohammad Ghavamzadeh, Sridhar Mahadevan

Browse the full ICML paper archive.