Improving the Price of Anarchy via Predictions in Parallel-Link Networks

2025-07-10

AI Summary

This paper addresses social cost degradation in parallel-link nonatomic congestion games with affine cost functions, arising from selfish routing, and investigates how machine learning–based prediction can inform the design of robust coordination mechanisms. Method: We propose a class of consistency mechanisms grounded in simple arrival-rate predictions. Leveraging price-of-anarchy analysis and error-tolerant mechanism design, we integrate predictive signals while guaranteeing robustness against prediction inaccuracies. Contribution/Results: We provide the first complete characterization of such mechanisms under all monotone cost functions. The mechanisms are error-adaptive: they achieve social optimality under perfect prediction, and their approximation ratio degrades smoothly as prediction error increases—while preserving theoretically optimal robustness within a pre-specified error tolerance. This framework achieves the optimal trade-off between predictive performance and worst-case robustness, advancing the theory of learning-augmented algorithmic game design.

Technology Category

• Multiagent Systems: Mechanism Design
• Game Theory and Economic Paradigms: Mechanism Design
• Machine Learning: Online Learning & Bandits

Application Category

• Economics, Online Markets and Human Computation: Incentives in network design for Web infrastructures and ecosystems
• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Graph Algorithms and Modeling for the Web: Algorithms and analysis for incomplete, noisy, or partially observed Web-related graphs

Abstract

We study non-atomic congestion games on parallel-link networks with affine cost functions. We investigate the power of machine-learned predictions in the design of coordination mechanisms aimed at minimizing the impact of selfishness. Our main results demonstrate that enhancing coordination mechanisms with a simple advice on the input rate can optimize the social cost whenever the advice is accurate (consistency), while only incurring minimal losses even when the predictions are arbitrarily inaccurate (bounded robustness). Moreover, we provide a full characterization of the consistent mechanisms that holds for all monotone cost functions, and show that our suggested mechanism is optimal with respect to the robustness. We further explore the notion of smoothness within this context: we extend our mechanism to achieve error-tolerance, i.e. we provide an approximation guarantee that degrades smoothly as a function of the prediction error, up to a predetermined threshold, while achieving a bounded robustness.

Problem

• Optimizing social cost in congestion games using predictions
• Characterizing consistent mechanisms for monotone cost functions
• Achieving error-tolerance with smooth degradation in performance

Innovation

• Enhancing coordination mechanisms with input rate advice
• Characterizing consistent mechanisms for monotone costs
• Extending mechanism for error-tolerant smoothness guarantees