MACOptions: Multi-Agent Learning with Centralized Controller and Options Framework

TL;DR

This paper explores multi-agent planning using a centralized controller, comparing different learning approaches including Q-learning with Options, and demonstrates the effectiveness of planning in multi-agent systems.

Contribution

It introduces a framework combining centralized control with options-based Q-learning for improved multi-agent planning.

Findings

Q-learning with Options outperforms simple Q-learning and random policies.

Centralized planning enhances multi-agent coordination.

Planner integration significantly improves performance.

Abstract

These days automation is being applied everywhere. In every environment, planning for the actions to be taken by the agents is an important aspect. In this paper, we plan to implement planning for multi-agents with a centralized controller. We compare three approaches: random policy, Q-learning, and Q-learning with Options Framework. We also show the effectiveness of planners by showing performance comparison between Q-Learning with Planner and without Planner.