Towards Assumption-free Bias Mitigation

TL;DR

This paper introduces an assumption-free framework that automatically detects and mitigates bias in machine learning models by modeling feature interactions, addressing privacy concerns and limitations of previous methods.

Contribution

The work presents a novel bias mitigation framework that does not rely on sensitive or manually defined attributes, using feature interaction modeling to identify and reduce bias.

Findings

Significantly reduces unfair prediction behaviors on real-world datasets.

Effectively detects biased feature interactions without sensitive attribute data.

Outperforms existing bias mitigation methods in experimental evaluations.

Abstract

Despite the impressive prediction ability, machine learning models show discrimination towards certain demographics and suffer from unfair prediction behaviors. To alleviate the discrimination, extensive studies focus on eliminating the unequal distribution of sensitive attributes via multiple approaches. However, due to privacy concerns, sensitive attributes are often either unavailable or missing in real-world scenarios. Therefore, several existing works alleviate the bias without sensitive attributes. Those studies face challenges, either in inaccurate predictions of sensitive attributes or the need to mitigate unequal distribution of manually defined non-sensitive attributes related to bias. The latter requires strong assumptions about the correlation between sensitive and non-sensitive attributes. As data distribution and task goals vary, the strong assumption on non-sensitive attributes may not be valid and require domain expertise. In this work, we propose an assumption-free framework to detect the related attributes automatically by modeling feature interaction for bias mitigation. The proposed framework aims to mitigate the unfair impact of identified biased feature interactions. Experimental results on four real-world datasets demonstrate that our proposed framework can significantly alleviate unfair prediction behaviors by considering biased feature interactions.