Multi-Pass Q-Networks for Deep Reinforcement Learning with Parameterised Action Spaces
This addresses a specific bottleneck in reinforcement learning for domains requiring combined discrete and continuous actions, representing an incremental improvement over prior methods.
The paper tackles the problem of reinforcement learning with parameterised action spaces, where P-DQN's approach invalidates theoretical foundations, and proposes MP-DQN, which significantly outperforms P-DQN and other algorithms in data efficiency and converged policy performance on domains like Platform, Robot Soccer Goal, and Half Field Offense.
Parameterised actions in reinforcement learning are composed of discrete actions with continuous action-parameters. This provides a framework for solving complex domains that require combining high-level actions with flexible control. The recent P-DQN algorithm extends deep Q-networks to learn over such action spaces. However, it treats all action-parameters as a single joint input to the Q-network, invalidating its theoretical foundations. We analyse the issues with this approach and propose a novel method, multi-pass deep Q-networks, or MP-DQN, to address them. We empirically demonstrate that MP-DQN significantly outperforms P-DQN and other previous algorithms in terms of data efficiency and converged policy performance on the Platform, Robot Soccer Goal, and Half Field Offense domains.