Toward a better trade-off between performance and fairness with kernel-based distribution matching
This work addresses fairness in machine learning for real-world deployment, though it appears incremental as it builds on existing regularization approaches.
The paper tackles the problem of balancing classifier performance and fairness by introducing a MinDiff framework that uses kernel-based distribution matching to regularize classifiers toward different fairness metrics, achieving real-world improvements in two large-scale industrial systems.
As recent literature has demonstrated how classifiers often carry unintended biases toward some subgroups, deploying machine learned models to users demands careful consideration of the social consequences. How should we address this problem in a real-world system? How should we balance core performance and fairness metrics? In this paper, we introduce a MinDiff framework for regularizing classifiers toward different fairness metrics and analyze a technique with kernel-based statistical dependency tests. We run a thorough study on an academic dataset to compare the Pareto frontier achieved by different regularization approaches, and apply our kernel-based method to two large-scale industrial systems demonstrating real-world improvements.