Hierarchical Deep Multiagent Reinforcement Learning with Temporal Abstraction

TL;DR

This paper introduces hierarchical deep multiagent reinforcement learning with temporal abstraction to improve coordination in environments with sparse, delayed rewards, demonstrating effectiveness in multiagent trash collection and a mobile game.

Contribution

It proposes three hierarchical deep MARL architectures and a new experience replay mechanism to handle sparse rewards and non-stationarity in multiagent settings.

Findings

Hierarchical architectures outperform baseline methods.

The new replay mechanism improves learning efficiency.

Effective in multiagent trash collection and Fever Basketball Defense.

Abstract

Multiagent reinforcement learning (MARL) is commonly considered to suffer from non-stationary environments and exponentially increasing policy space. It would be even more challenging when rewards are sparse and delayed over long trajectories. In this paper, we study hierarchical deep MARL in cooperative multiagent problems with sparse and delayed reward. With temporal abstraction, we decompose the problem into a hierarchy of different time scales and investigate how agents can learn high-level coordination based on the independent skills learned at the low level. Three hierarchical deep MARL architectures are proposed to learn hierarchical policies under different MARL paradigms. Besides, we propose a new experience replay mechanism to alleviate the issue of the sparse transitions at the high level of abstraction and the non-stationarity of multiagent learning. We empirically demonstrate the effectiveness of our approaches in two domains with extremely sparse feedback: (1) a variety of Multiagent Trash Collection tasks, and (2) a challenging online mobile game, i.e., Fever Basketball Defense.