Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2023 Mar 28:1-32.
doi: 10.1007/s10994-023-06319-8. Online ahead of print.

PreCoF: counterfactual explanations for fairness

Sofie Goethals  1 , David Martens  1 , Toon Calders  2
Affiliations

PreCoF: counterfactual explanations for fairness

Sofie Goethals et al. Mach Learn. .

Abstract

This paper studies how counterfactual explanations can be used to assess the fairness of a model. Using machine learning for high-stakes decisions is a threat to fairness as these models can amplify bias present in the dataset, and there is no consensus on a universal metric to detect this. The appropriate metric and method to tackle the bias in a dataset will be case-dependent, and it requires insight into the nature of the bias first. We aim to provide this insight by integrating explainable AI (XAI) research with the fairness domain. More specifically, apart from being able to use (Predictive) Counterfactual Explanations to detect explicit bias when the model is directly using the sensitive attribute, we show that it can also be used to detect implicit bias when the model does not use the sensitive attribute directly but does use other correlated attributes leading to a substantial disadvantage for a protected group. We call this metric PreCoF, or Predictive Counterfactual Fairness. Our experimental results show that our metric succeeds in detecting occurrences of implicit bias in the model by assessing which attributes are more present in the explanations of the protected group compared to the unprotected group. These results could help policymakers decide on whether this discrimination is justified or not.

Keywords: Counterfactual explanations; Data science ethics; Explainable Artificial Intelligence; Fairness.

PubMed Disclaimer

Conflict of interest statement

Conflict of interestNot applicable.

Figures

Fig. 1
Fig. 1
Difference in PreCoF for men and women in the Adult Income dataset
Fig. 2
Fig. 2
Relationship between sex and the attributes marital status/relationship
Fig. 3
Fig. 3
Difference in PreCoF for foreigners and locals in the Catalonia Juvenile dataset
Fig. 4
Fig. 4
Catalonia Juvenile dataset: analysis
Fig. 5
Fig. 5
Crime and Communities dataset: analysis
Fig. 6
Fig. 6
Difference in explanations for boys and girls in the Student performance dataset
Fig. 7
Fig. 7
Student performance dataset: analysis
Fig. 8
Fig. 8
Law Admission dataset: analysis
Fig. 9
Fig. 9
PreSHAPF
Fig. 10
Fig. 10
Additional illustration with a transparent machine learning model to show the difference between PreCoF and PreSHAPF
See this image and copyright information in PMC

References

    1. Adadi A, Berrada M. Peeking inside the black-box: A survey on explainable artificial intelligence (XAI) IEEE Access. 2018;6:52138–52160. doi: 10.1109/ACCESS.2018.2870052. - DOI
    1. Asuncion, A., & Newman, D. (2007). UCI Machine Learning Repository.
    1. Black, E., Yeom, S., & Fredrikson, M. (2020). Fliptest: Fairness testing via optimal transport. In: Proceedings of the 2020 conference on fairness, accountability, and transparency (pp. 111–121).
    1. Bonchi F, Hajian S, Mishra B, Ramazzotti D. Exposing the probabilistic causal structure of discrimination. International Journal of Data Science and Analytics. 2017;3(1):1–21. doi: 10.1007/s41060-016-0040-z. - DOI
    1. Bordt, S., Finck, M., Raidl, E., & von Luxburg, U. (2022). Post-hoc explanations fail to achieve their purpose in adversarial contexts. arXiv preprint arXiv:2201.10295 .

LinkOut - more resources