REValueD: Regularised Ensemble Value-Decomposition for Factorisable Markov Decision Processes
This addresses a bottleneck in reinforcement learning for robotics and control tasks, offering an incremental improvement over existing value-decomposition methods.
The paper tackled the challenge of high-dimensional discrete action spaces in reinforcement learning by introducing REValueD, a regularised ensemble value-decomposition algorithm, which showed superior performance on discretised DeepMind Control Suite tasks, particularly in humanoid and dog tasks.
Discrete-action reinforcement learning algorithms often falter in tasks with high-dimensional discrete action spaces due to the vast number of possible actions. A recent advancement leverages value-decomposition, a concept from multi-agent reinforcement learning, to tackle this challenge. This study delves deep into the effects of this value-decomposition, revealing that whilst it curtails the over-estimation bias inherent to Q-learning algorithms, it amplifies target variance. To counteract this, we present an ensemble of critics to mitigate target variance. Moreover, we introduce a regularisation loss that helps to mitigate the effects that exploratory actions in one dimension can have on the value of optimal actions in other dimensions. Our novel algorithm, REValueD, tested on discretised versions of the DeepMind Control Suite tasks, showcases superior performance, especially in the challenging humanoid and dog tasks. We further dissect the factors influencing REValueD's performance, evaluating the significance of the regularisation loss and the scalability of REValueD with increasing sub-actions per dimension.