This paper studies the challenging continual learning (CL) setting of Cl...
As AI agents are increasingly used in the real open world with unknowns ...
Strategy card game is a well-known genre that is demanding on the intell...
Deep Reinforcement Learning combined with Fictitious Play shows impressi...
Continual learning (CL) learns a sequence of tasks incrementally. There ...
To obtain higher sample efficiency and superior final performance
simult...
Existing continual learning techniques focus on either task incremental
...
Deep Q Network (DQN) firstly kicked the door of deep reinforcement learn...
Policy-based reinforcement learning methods suffer from the policy colla...
Building on the breakthrough of reinforcement learning, this paper intro...