# Oblivious Data for Fairness with Kernels

arXiv: https://arxiv.org/abs/2002.02901 · HTML: https://mlsanity.com/paper/2002.02901/

## Key facts

- Cited by 5 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 7 Feb 2020.

## Built on by (most cited first, 5 of 5)

- [FARE: Provably Fair Representation Learning with Practical Certificates](https://mlsanity.com/paper/2210.07213/) (2022)
- [Mean Parity Fair Regression in RKHS](https://mlsanity.com/paper/2302.10409/) (2023)
- [Should Bank Stress Tests Be Fair?](https://mlsanity.com/paper/2207.13319/) (2022)
- [Lightweight Fairness for LLM-Based Recommendations via Kernelized Projection and Gated Adapters](https://mlsanity.com/paper/2603.23780/) (2026)
- [Extending Fair Null-Space Projections for Continuous Attributes to Kernel Methods](https://mlsanity.com/paper/2511.03304/) (2025)

## Abstract (arXiv)

We investigate the problem of algorithmic fairness in the case where sensitive and non-sensitive features are available and one aims to generate new, `oblivious', features that closely approximate the non-sensitive features, and are only minimally dependent on the sensitive ones. We study this question in the context of kernel methods. We analyze a relaxed version of the Maximum Mean Discrepancy criterion which does not guarantee full independence but makes the optimization problem tractable. We derive a closed-form solution for this relaxed optimization problem and complement the result with a study of the dependencies between the newly generated features and the sensitive ones. Our key ingredient for generating such oblivious features is a Hilbert-space-valued conditional expectation, which needs to be estimated from data. We propose a plug-in approach and demonstrate how the estimation errors can be controlled. While our techniques help reduce the bias, we would like to point out that no post-processing of any dataset could possibly serve as an alternative to well-designed experiments.