Risk-averse Fair Multi-class Classification

2025-09-06

AI Summary

This paper addresses multi-class classification under high noise, limited samples, and unreliable labels. We propose a novel risk-averse and fair classification framework. Methodologically, we pioneer the integration of systemic risk theory and coherent risk measures into multi-class modeling, formulating a two-stage stochastic program; it is solved via a nonlinear aggregation mechanism and a risk-averse regularization decomposition algorithm, augmented with group fairness constraints. Our contributions are: (1) extending coherent risk measures to multi-class settings, moving beyond conventional expected risk minimization; and (2) jointly enhancing robustness and fairness, yielding substantial improvements in generalization under label noise and data scarcity. Experiments demonstrate that our method outperforms state-of-the-art approaches on unseen data, with performance gains amplifying as the number of classes increases. Moreover, all evaluated fairness metrics—e.g., equalized odds, demographic parity—show significant improvement.

Technology Category

• Machine Learning: Multi-class/Multi-label Learning & Extreme Classification
• Reasoning under Uncertainty: Stochastic Optimization
• 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
• Responsible Web: Machine-in-the-loop, human agency and autonomy

Abstract

We develop a new classification framework based on the theory of coherent risk measures and systemic risk. The proposed approach is suitable for multi-class problems when the data is noisy, scarce (relative to the dimension of the problem), and the labeling might be unreliable. In the first part of our paper, we provide the foundation of the use of systemic risk models and show how to apply it in the context of linear and kernel-based multi-class problems. More advanced formulation via a system-theoretic approach with non-linear aggregation is proposed, which leads to a two-stage stochastic programming problem. A risk-averse regularized decomposition method is designed to solve the problem. We use a popular multi-class method as a benchmark in the performance analysis of the proposed classification methods. We illustrate our ideas by proposing several generalization of that method by the use of coherent measures of risk. The viability of the proposed risk-averse methods are supported theoretically and numerically. Additionally, we demonstrate that the application of systemic risk measures facilitates enforcing fairness in classification. Analysis and experiments regarding the fairness of the proposed models are carefully conducted. For all methods, our numerical experiments demonstrate that they are robust in the presence of unreliable training data and perform better on unknown data than the methods minimizing expected classification errors. Furthermore, the performance improves when the number of classes increases.

Problem

• Develops risk-averse multi-class classification for noisy, scarce data
• Addresses unreliable labeling through coherent risk measure framework
• Enforces classification fairness using systemic risk measures

Innovation

• Systemic risk measures for robust multi-class classification
• Risk-averse regularized decomposition method for optimization
• Non-linear aggregation via system-theoretic stochastic programming