Learning Certified Individually Fair Representations

Abstract

Fair representation learning provides an effective way of enforcing fairness constraints without compromising utility for downstream users. A desirable family of such fairness constraints, each requiring similar treatment for similar individuals, is known as individual fairness. In this work, we introduce the first method that enables data consumers to obtain certificates of individual fairness for existing and new data points. The key idea is to map similar individuals to close latent representations and leverage this latent proximity to certify individual fairness. That is, our method enables the data producer to learn and certify a representation where for a data point all similar individuals are at ${\ell }_{\infty }$-distance at most $\epsilon$, thus allowing data consumers to certify individual fairness by proving $\epsilon$-robustness of their classifier. Our experimental evaluation on five real-world datasets and several fairness constraints demonstrates the expressivity and scalability of our approach.

arXiv commentConference Paper at NeurIPS 2020

References in corpus (23)

• PyTorch: An Imperative Style, High-Performance Deep Learning Library
• Equality of Opportunity in Supervised Learning
• Man is to Computer Programmer as Woman is to Homemaker? Debiasing Word Embeddings
• Counterfactual Fairness
• Preventing Fairness Gerrymandering: Auditing and Learning for Subgroup Fairness
• Efficient Neural Network Robustness Certification with General Activation Functions
• Evaluating Robustness of Neural Networks with Mixed Integer Programming
• The Frontiers of Fairness in Machine Learning
• Flexibly Fair Representation Learning by Disentanglement
• A Reductions Approach to Fair Classification
• Fairness Under Composition
• Operationalizing Individual Fairness with Pairwise Fair Representations
• Learning Controllable Fair Representations
• Training individually fair ML models with Sensitive Subspace Robustness
• Provably Fair Representations
• Individual Fairness for $k$-Clustering
• Two Simple Ways to Learn Individual Fairness Metrics from Data
• An Empirical Study on Learning Fairness Metrics for COMPAS Data with Human Supervision
• A Spectral View of Adversarially Robust Features
• A Center in Your Neighborhood: Fairness in Facility Location
• Learning Adversarially Robust Representations via Worst-Case Mutual Information Maximization
• Perfectly Parallel Fairness Certification of Neural Networks
• Learning Individually Fair Classifier with Path-Specific Causal-Effect Constraint

Cited by in corpus (3)

• PRIMA: General and Precise Neural Network Certification via Scalable Convex Hull Approximations
• Toward Operationalizing Pipeline-aware ML Fairness: A Research Agenda for Developing Practical Guidelines and Tools
• TableGAN-MCA: Evaluating Membership Collisions of GAN-Synthesized Tabular Data Releasing