Hierarchical Orchestra of Policies

Thomas P Cannon, Özgür Simsek

Continual reinforcement learning poses a major challenge due to the tendency of agents to experience catastrophic forgetting when learning sequential tasks. In this paper, we intro…