Continuous Episodic Control
This addresses the limitation of episodic control methods to discrete actions, enabling faster learning for continuous control tasks, though it is an incremental extension.
The paper tackled the problem of applying episodic memory to continuous action spaces in reinforcement learning, introducing Continuous Episodic Control (CEC) which learns faster than state-of-the-art model-free and memory-augmented RL algorithms in sparse-reward environments.
Non-parametric episodic memory can be used to quickly latch onto high-rewarded experience in reinforcement learning tasks. In contrast to parametric deep reinforcement learning approaches in which reward signals need to be back-propagated slowly, these methods only need to discover the solution once, and may then repeatedly solve the task. However, episodic control solutions are stored in discrete tables, and this approach has so far only been applied to discrete action space problems. Therefore, this paper introduces Continuous Episodic Control (CEC), a novel non-parametric episodic memory algorithm for sequential decision making in problems with a continuous action space. Results on several sparse-reward continuous control environments show that our proposed method learns faster than state-of-the-art model-free RL and memory-augmented RL algorithms, while maintaining good long-run performance as well. In short, CEC can be a fast approach for learning in continuous control tasks.