Adaptive action supervision in reinforcement learning from real-world multi-agent demonstrations

TL;DR

This paper introduces an adaptive action supervision method in reinforcement learning that leverages real-world multi-agent demonstrations, effectively bridging the domain gap between real-world data and simulated environments.

Contribution

It proposes a novel approach combining RL and supervised learning using dynamic time warping to select actions, enhancing reproducibility and generalization in multi-agent RL tasks.

Findings

Achieved a balance between imitation and reward-based learning in experiments.

Successfully applied the method to football tracking data with large domain gaps.

Demonstrated improved performance over baseline methods.

Abstract

Modeling of real-world biological multi-agents is a fundamental problem in various scientific and engineering fields. Reinforcement learning (RL) is a powerful framework to generate flexible and diverse behaviors in cyberspace; however, when modeling real-world biological multi-agents, there is a domain gap between behaviors in the source (i.e., real-world data) and the target (i.e., cyberspace for RL), and the source environment parameters are usually unknown. In this paper, we propose a method for adaptive action supervision in RL from real-world demonstrations in multi-agent scenarios. We adopt an approach that combines RL and supervised learning by selecting actions of demonstrations in RL based on the minimum distance of dynamic time warping for utilizing the information of the unknown source dynamics. This approach can be easily applied to many existing neural network architectures and provide us with an RL model balanced between reproducibility as imitation and generalization ability to obtain rewards in cyberspace. In the experiments, using chase-and-escape and football tasks with the different dynamics between the unknown source and target environments, we show that our approach achieved a balance between the reproducibility and the generalization ability compared with the baselines. In particular, we used the tracking data of professional football players as expert demonstrations in football and show successful performances despite the larger gap between behaviors in the source and target environments than the chase-and-escape task.