fairadapt: Causal Reasoning for Fair Data Pre-processing
This work addresses fairness in machine learning for applications involving sensitive attributes like gender and race, but it is incremental as it implements an existing causal method in a new package.
The authors tackled algorithmic bias by developing the R-package fairadapt, which uses causal inference for data pre-processing to address counterfactual questions like salary differences based on gender or race, aiming to eliminate discrimination and justify fair decisions.
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.