Improving Multi-agent Coordination by Learning to Estimate Contention

TL;DR

This paper introduces ALMA-Learning, a decentralized multi-agent learning algorithm that efficiently achieves near-optimal and fair coordination without inter-agent communication, suitable for large-scale systems and on-device deployment.

Contribution

The paper proposes ALMA-Learning, a novel multi-agent learning approach that overcomes traditional challenges by using ALMA heuristics for decentralized, communication-free coordination.

Findings

Achieves near-optimal (<5%) coordination loss in synthetic scenarios.

Ensures fair allocation among agents.

Demonstrates effectiveness in real-world meeting scheduling.

Abstract

We present a multi-agent learning algorithm, ALMA-Learning, for efficient and fair allocations in large-scale systems. We circumvent the traditional pitfalls of multi-agent learning (e.g., the moving target problem, the curse of dimensionality, or the need for mutually consistent actions) by relying on the ALMA heuristic as a coordination mechanism for each stage game. ALMA-Learning is decentralized, observes only own action/reward pairs, requires no inter-agent communication, and achieves near-optimal (<5% loss) and fair coordination in a variety of synthetic scenarios and a real-world meeting scheduling problem. The lightweight nature and fast learning constitute ALMA-Learning ideal for on-device deployment.