Fair Domain Generalization: An Information-Theoretic View

2025/07/08 by Tangzheng Lian, Lian, Tangzheng, Guanyu Hu +7 · 1 voice
Computer Science · Social Sciences · #Advanced Graph Neural Networks #Computer Vision and Pattern Recognition (cs.CV) #Domain Adaptation and Few-Shot Learning #Ethics and Social Impacts of AI #FOS: Computer and information sciences #Machine Learning (cs.LG)

paper · pdf · doi:10.48550/arxiv.2507.05823

openalex publication_date 2025/07/08 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

Domain generalization (DG) and algorithmic fairness are two critical challenges in machine learning. However, most DG methods focus only on minimizing expected risk in the unseen target domain without considering algorithmic fairness. Conversely, fairness methods typically do not account for domain shifts, so the fairness achieved during training may not generalize to unseen test domains. In this work, we bridge these gaps by studying the problem of Fair Domain Generalization (FairDG), which aims to minimize both expected risk and fairness violations in unseen target domains. We derive novel mutual information-based upper bounds for expected risk and fairness violations in multi-class classification tasks with multi-group sensitive attributes. These bounds provide key insights for algorithm design from an information-theoretic perspective. Guided by these insights, we introduce PAFDG (Pareto-Optimal Fairness for Domain Generalization), a practical framework that solves the FairDG problem and models the utility-fairness trade-off through Pareto optimization. Experiments on real-world vision and language datasets show that PAFDG achieves superior utility-fairness trade-offs compared to existing methods.

Citations

• FEED: Fairness-Enhanced Meta-Learning for Domain Generalization
• Rethinking Affect Analysis: A Protocol for Ensuring Fairness and Consistency
• Algorithmic Fairness Generalization under Covariate and Dependence Shifts Simultaneously
• Towards Counterfactual Fairness-aware Domain Generalization in Changing Environments
• Diversify Your Vision Datasets with Automatic Diffusion-Based Augmentation
• Fairness and Accuracy under Domain Generalization
• How Robust is Your Fairness? Evaluating and Sustaining Fairness under Unseen Distribution Shifts
• Fairness Transferability Subject to Bounded Distribution Shift
• Towards Principled Disentanglement for Domain Generalization
• On Characterizing the Trade-off in Invariant Representation Learning
• Robust Fairness under Covariate Shift
• Learning De-biased Representations with Biased Representations
• A Survey on Bias and Fairness in Machine Learning
• Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks
• On Variational Bounds of Mutual Information
• Mitigating Information Leakage in Image Representations: A Maximum\n Entropy Approach
• Learning Not to Learn: Training Deep Neural Networks with Biased Data
• Learning to Generalize: Meta-Learning for Domain Generalization
• Equality of Opportunity in Supervised Learning
• Deep CORAL: Correlation Alignment for Deep Domain Adaptation
• Deep Residual Learning for Image Recognition
• Deep Learning Face Attributes in the Wild
• Measuring and testing dependence by correlation of distances
• Swin Transformer: Hierarchical Vision Transformer using Shifted Windows
• Swin Transformer: Hierarchical Vision Transformer using Shifted Windows

Discussions

• Our paper “Fair Domain Generalization: An Information-Theoretic View” is just selected to be presented at #AAAI2026 as ORAL! Among 31,000 abstract submissions with an acceptance rate of ~17% 😅🦾🧠 Pr [bsky, 7 points, 2 comments]

Related

• Towards Counterfactual Fairness-aware Domain Generalization in Changing Environments
• FEED: Fairness-Enhanced Meta-Learning for Domain Generalization
• Fairness and Accuracy under Domain Generalization
• Algorithmic Fairness Generalization under Covariate and Dependence Shifts Simultaneously
• FAIRM: Learning invariant representations for algorithmic fairness and domain generalization with minimax optimality