Open the Black Box: Step-based Policy Updates for Temporally-Correlated Episodic Reinforcement Learning
This addresses the challenge of implementing smooth and efficient policies on real hardware for reinforcement learning practitioners, though it appears incremental as it builds on existing episodic and step-based methods.
The paper tackled the problem of inefficient exploration and unsmooth trajectories in reinforcement learning by introducing Temporally-Correlated Episodic RL (TCE), which combines step-based and episodic approaches to achieve comparable performance to episodic methods while maintaining data efficiency similar to state-of-the-art step-based RL.
Current advancements in reinforcement learning (RL) have predominantly focused on learning step-based policies that generate actions for each perceived state. While these methods efficiently leverage step information from environmental interaction, they often ignore the temporal correlation between actions, resulting in inefficient exploration and unsmooth trajectories that are challenging to implement on real hardware. Episodic RL (ERL) seeks to overcome these challenges by exploring in parameters space that capture the correlation of actions. However, these approaches typically compromise data efficiency, as they treat trajectories as opaque \emph{black boxes}. In this work, we introduce a novel ERL algorithm, Temporally-Correlated Episodic RL (TCE), which effectively utilizes step information in episodic policy updates, opening the 'black box' in existing ERL methods while retaining the smooth and consistent exploration in parameter space. TCE synergistically combines the advantages of step-based and episodic RL, achieving comparable performance to recent ERL methods while maintaining data efficiency akin to state-of-the-art (SoTA) step-based RL.