Generalising Discrete Action Spaces with Conditional Action Trees
This addresses the problem of adapting RL algorithms to diverse action space formats, particularly for tasks with complex action spaces, though it appears incremental in its approach.
The paper tackles the lack of standardization in reinforcement learning action spaces by introducing Conditional Action Trees, which structure and reduce action spaces through decomposition, validated in experiments from basic discrete to large combinatorial environments.
There are relatively few conventions followed in reinforcement learning (RL) environments to structure the action spaces. As a consequence the application of RL algorithms to tasks with large action spaces with multiple components require additional effort to adjust to different formats. In this paper we introduce {\em Conditional Action Trees} with two main objectives: (1) as a method of structuring action spaces in RL to generalise across several action space specifications, and (2) to formalise a process to significantly reduce the action space by decomposing it into multiple sub-spaces, favoring a multi-staged decision making approach. We show several proof-of-concept experiments validating our scheme, ranging from environments with basic discrete action spaces to those with large combinatorial action spaces commonly found in RTS-style games.