Provably Fair Representations

Abstract

Machine learning systems are increasingly used to make decisions about people's lives, such as whether to give someone a loan or whether to interview someone for a job. This has led to considerable interest in making such machine learning systems fair. One approach is to transform the input data used by the algorithm. This can be achieved by passing each input data point through a representation function prior to its use in training or testing. Techniques for learning such representation functions from data have been successful empirically, but typically lack theoretical fairness guarantees. We show that it is possible to prove that a representation function is fair according to common measures of both group and individual fairness, as well as useful with respect to a target task. These provable properties can be used in a governance model involving a data producer, a data user and a data regulator, where there is a separation of concerns between fairness and target task utility to ensure transparency and prevent perverse incentives. We formally define the 'cost of mistrust' of using this model compared to the setting where there is a single trusted party, and provide bounds on this cost in particular cases. We present a practical approach to learning fair representation functions and apply it to financial and criminal justice datasets. We evaluate the fairness and utility of these representation functions using measures motivated by our theoretical results.

References in corpus (4)

• Data Decisions and Theoretical Implications when Adversarially Learning Fair Representations
• A survey on measuring indirect discrimination in machine learning
• Optimized Data Pre-Processing for Discrimination Prevention
• The cost of fairness in classification

Cited by in corpus (15)

• Fairness in Machine Learning
• The Frontiers of Fairness in Machine Learning
• A Clarification of the Nuances in the Fairness Metrics Landscape
• Learning Adversarially Fair and Transferable Representations
• Learning Certified Individually Fair Representations
• Fair Representation: Guaranteeing Approximate Multiple Group Fairness for Unknown Tasks
• Removing biased data to improve fairness and accuracy
• Learning Fair and Transferable Representations
• On the Fairness of Disentangled Representations
• Hierarchical VampPrior Variational Fair Auto-Encoder
• Removing Spurious Features can Hurt Accuracy and Affect Groups Disproportionately
• Ethical Quantum Computing: A Roadmap
• Computability, Complexity, Consistency and Controllability: A Four C's Framework for cross-disciplinary Ethical Algorithm Research
• A Distributed Fair Machine Learning Framework with Private Demographic Data Protection
• Fair Kernel Regression via Fair Feature Embedding in Kernel Space