Utilitarian Distortion with Predictions

2025-02-10

AI Summary

This paper investigates the consistency–robustness trade-off in prediction-augmented social choice mechanisms, focusing on single-winner voting and one-sided matching. Given agents’ ordinal preferences and access to noisy cardinal utility predictions, we design mechanisms that improve performance when predictions are accurate while guaranteeing worst-case distortion bounds when predictions fail. We establish the first tight theoretical characterization of this trade-off, proving that our mechanisms strictly improve upon the optimal distortion achievable by purely ordinal mechanisms. Leveraging tools from combinatorial optimization, game theory, and prediction-augmented algorithm design, we provide both the first optimal theoretical foundation for learning-enhanced ordinal mechanisms and constructive procedures for their implementation.

Technology Category

• Game Theory and Economic Paradigms: Social Choice / Voting
• Multiagent Systems: Mechanism Design
• Machine Learning: Learning Preferences or Rankings

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Economics, Online Markets and Human Computation: Social networks and social learning
• Social Networks and Social Media: Influence propagation, information diffusion, and the prediction on networks

Abstract

We study the utilitarian distortion of social choice mechanisms under the recently proposed learning-augmented framework where some (possibly unreliable) predicted information about the preferences of the agents is given as input. In particular, we consider two fundamental social choice problems: single-winner voting and one-sided matching. In these settings, the ordinal preferences of the agents over the alternatives (either candidates or items) is known, and some prediction about their underlying cardinal values is also provided. The goal is to leverage the prediction to achieve improved distortion guarantees when it is accurate, while simultaneously still achieving reasonable worst-case bounds when it is not. This leads to the notions of consistency and robustness, and the quest to achieve the best possible tradeoffs between the two. We show tight tradeoffs between the consistency and robustness of ordinal mechanisms for single-winner voting and one-sided matching, for different levels of information provided by as prediction.

Problem

• Study utilitarian distortion in social choice mechanisms.
• Leverage predictions to improve distortion guarantees.
• Achieve tradeoffs between consistency and robustness.

Innovation

• learning-augmented social choice
• tight consistency-robustness tradeoffs
• ordinal mechanisms with predictions