fairadapt: Causal Reasoning for Fair Data Pre-processing

TL;DR

Fairadapt is an R-package that uses causal inference to pre-process data, enabling counterfactual reasoning to mitigate bias related to sensitive attributes like gender and race in machine learning.

Contribution

It introduces a causal graphical model-based pre-processing method for fair data treatment, allowing individual-level counterfactual analysis to reduce discrimination.

Findings

Enables counterfactual reasoning for fairness.

Addresses hypothetical 'what-if' scenarios.

Provides relaxations for causal pathways.

Abstract

Machine learning algorithms are useful for various predictions tasks, but they can also learn how to discriminate, based on gender, race or other sensitive attributes. This realization gave rise to the field of fair machine learning, which aims to measure and mitigate such algorithmic bias. This manuscript describes the R-package fairadapt, which implements a causal inference pre-processing method. By making use of a causal graphical model and the observed data, the method can be used to address hypothetical questions of the form "What would my salary have been, had I been of a different gender/race?". Such individual level counterfactual reasoning can help eliminate discrimination and help justify fair decisions. We also discuss appropriate relaxations which assume certain causal pathways from the sensitive attribute to the outcome are not discriminatory.