Toward a better trade-off between performance and fairness with kernel-based distribution matching
By Flavien Prost, Hai Qian, Qiuwen Chen, Ed H. Chi, Jilin Chen, Alex Beutel. 2019
49 citations recorded by the source corpus.
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.
Subfields: Safety Research
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Safety Research
- Nithum Thain
- Alex Beutel
- Ed H. Chi
- Jilin Chen
- Flavien Prost
- Silvia Chiappa
- Margaret Mitchell
- Luca Oneto
Citations in the corpus
- Fairness without Demographics through Adversarially Reweighted Learning (2020, methodology): “…standard • DRO [17]: a group-agnostic distributionally robust optimization • IPW[22]:A group-aware commonre-weightingapproach,which • Min-Diff [41]: A group-aware group-fairness approach that aims Setup and Parameter Tuning: We use the same experimental As our proposed ARL model has…”