EASpace: Enhanced Action Space for Policy Transfer

TL;DR

EASpace introduces a novel macro action formulation that accelerates policy learning by integrating multiple expert policies and an intrinsic reward, improving exploration and data efficiency in complex tasks.

Contribution

The paper proposes EASpace, a new macro action formulation that enhances policy transfer by integrating multiple expert policies with an intrinsic reward mechanism.

Findings

EASpace accelerates learning in grid-based and pursuit tasks.

Theoretical convergence of the learning rule is established.

EASpace is effective in physical system implementations.

Abstract

Formulating expert policies as macro actions promises to alleviate the long-horizon issue via structured exploration and efficient credit assignment. However, traditional option-based multi-policy transfer methods suffer from inefficient exploration of macro action's length and insufficient exploitation of useful long-duration macro actions. In this paper, a novel algorithm named EASpace (Enhanced Action Space) is proposed, which formulates macro actions in an alternative form to accelerate the learning process using multiple available sub-optimal expert policies. Specifically, EASpace formulates each expert policy into multiple macro actions with different execution {times}. All the macro actions are then integrated into the primitive action space directly. An intrinsic reward, which is proportional to the execution time of macro actions, is introduced to encourage the exploitation of useful macro actions. The corresponding learning rule that is similar to Intra-option Q-learning is employed to improve the data efficiency. Theoretical analysis is presented to show the convergence of the proposed learning rule. The efficiency of EASpace is illustrated by a grid-based game and a multi-agent pursuit problem. The proposed algorithm is also implemented in physical systems to validate its effectiveness.