Learning-Augmented Coordination Mechanisms

概要

The inefficiency of decentralized resource allocation is classically measured by the Price of Anarchy (PoA) and the Price of Stability (PoS). The most widely studied and foundational class of these models are atomic and non-atomic congestion games, for which a robust theory quantifying the inefficiency of equilibria is well-established. To mitigate the effects of selfish behavior, we employ Coordination Mechanisms, which modify resource costs to incentivize socially improved equilibria. However, standard Coordination Mechanisms assume unknown, adversarial demand, restricting positive results mainly to specialized networks such as parallel links. We address this limitation by studying learning-augmented coordination mechanisms, where the mechanism is endowed with a potentially inaccurate prediction of the demand. We study general atomic and non-atomic congestion games with polynomial latency functions of degree $d$. Our main positive result is a learning-augmented coordination mechanism tuned by a confidence parameter $β$ which achieves a "best of both worlds" type of result: approximately optimal performance for an accurate prediction without sacrificing the worst-case guarantee of a good equilibrium (PoS). We further establish bounds on PoA-Robustness (worst equilibrium performance under cost modification). Our construction relies on transforming approximate equilibria of the original game into exact equilibria of the modified game using an approximate potential function. We complement our positive results with impossibility results. We first show that requiring all equilibria to have optimal social cost under an accurate prediction precludes bounded robustness. We also show a strong error tolerance limit in non-atomic games: even infinitesimal prediction errors can cause the performance of a mechanism with optimal consistency to revert to the original PoA.

参考リンク

• Paper ↗

リンク先のライセンス・再現性・データ条件をご確認ください。自動分類や実装の目安は保証ではありません。

実装を始めるプロンプト

参考リンクを確認したうえで、次の内容をコピーして利用できます。

# 論文・記事
Learning-Augmented Coordination Mechanisms

# 概要
The inefficiency of decentralized resource allocation is classically measured by the Price of Anarchy (PoA) and the Price of Stability (PoS). The most widely studied and foundational class of these…

# 原典
https://arxiv.org/abs/2610.36508

上記の手法を実装してください。実行前に依存関係とライセンスを確認してください。