Individual Fairness In Strategic Classification

2026-02-04

AI Summary

This work proposes the first provably individually fair randomized threshold classifier for strategic classification, where individuals may manipulate their features to influence model decisions, thereby undermining individual fairness. By formulating and solving a linear program to derive an optimal randomized classification policy, the method simultaneously guarantees both individual and group fairness under strategic behavior. The approach innovatively integrates the theoretical framework of individual fairness with explicit modeling of strategic manipulation. Empirical evaluations on real-world datasets demonstrate that the proposed classifier effectively mitigates unfairness while maintaining high predictive accuracy, thus achieving a favorable trade-off between fairness and performance in strategic settings.

Technology Category

• Game Theory and Economic Paradigms: Fair Division
• Machine Learning: Ethics, Bias, and Fairness
• Computer Vision: Bias, Fairness & Privacy

Application Category

• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Economics, Online Markets and Human Computation: Fairness, privacy, and diversity in economic environments
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Strategic classification, where individuals modify their features to influence machine learning (ML) decisions, presents critical fairness challenges. While group fairness in this setting has been widely studied, individual fairness remains underexplored. We analyze threshold-based classifiers and prove that deterministic thresholds violate individual fairness. Then, we investigate the possibility of using a randomized classifier to achieve individual fairness. We introduce conditions under which a randomized classifier ensures individual fairness and leverage these conditions to find an optimal and individually fair randomized classifier through a linear programming problem. Additionally, we demonstrate that our approach can be extended to group fairness notions. Experiments on real-world datasets confirm that our method effectively mitigates unfairness and improves the fairness-accuracy trade-off.

Problem

• individual fairness
• strategic classification
• fairness
• machine learning
• threshold-based classifiers

Innovation

• individual fairness
• strategic classification
• randomized classifier
• linear programming
• fairness-accuracy trade-off