Please use this identifier to cite or link to this item:
|Publication type:||Conference paper|
|Type of review:||Peer review (publication)|
|Title:||Enforcing group fairness in algorithmic decision making : utility maximization under sufficiency|
|Proceedings:||Proceedings of the 2022 ACM Conference on Fairness, Accountability, and Transparency|
|Conference details:||5th ACM Conference on Fairness, Accountability, and Transparency (FAccT), Seoul, Republic of Korea, 21-24 June 2022|
|Publisher / Ed. Institution:||Association for Computing Machinery|
|Publisher / Ed. Institution:||New York|
|Subjects:||Computing methodology; Machine learning; Applied computing; Decision analysis; Group fairness metrics; Algorithmic fairness; Prediction-based decision making|
|Subject (DDC):||005: Computer programming, programs and data |
658.403: Decision making, information management
|Abstract:||Binary decision making classifiers are not fair by default. Fairness requirements are an additional element to the decision making rationale, which is typically driven by maximizing some utility function. In that sense, algorithmic fairness can be formulated as a constrained optimization problem. This paper contributes to the discussion on how to implement fairness, focusing on the fairness concepts of positive predictive value (PPV) parity, false omission rate (FOR) parity, and sufficiency (which combines the former two). We show that group-specific threshold rules are optimal for PPV parity and FOR parity, similar to well-known results for other group fairness criteria. However, depending on the underlying population distributions and the utility function, we find that sometimes an upper-bound threshold rule for one group is optimal: utility maximization under PPV parity (or FOR parity) might thus lead to selecting the individuals with the smallest utility for one group, instead of selecting the most promising individuals. This result is counter-intuitive and in contrast to the analogous solutions for statistical parity and equality of opportunity. We also provide a solution for the optimal decision rules satisfying the fairness constraint sufficiency. We show that more complex decision rules are required and that this leads to within-group unfairness for all but one of the groups. We illustrate our findings based on simulated and real data.|
|Fulltext version:||Accepted version|
|License (according to publishing contract):||Licence according to publishing contract|
|Departement:||School of Engineering|
|Organisational Unit:||Institute of Data Analysis and Process Design (IDP)|
|Published as part of the ZHAW project:||Algorithmic Fairness in data-based decision making: Combining ethics and technology|
|Appears in collections:||Publikationen School of Engineering|
Files in This Item:
|2022_Baumann-Hannak-Heitz_Enforcing-group-fairness-algorithmic-decision-making_ACM.pdf||Accepted Version||2.07 MB||Adobe PDF|
Show full item record
Baumann, J., Hannák, A., & Heitz, C. (2022). Enforcing group fairness in algorithmic decision making : utility maximization under sufficiency [Conference paper]. Proceedings of the 2022 ACM Conference on Fairness, Accountability, and Transparency, 2315–2326. https://doi.org/10.1145/3531146.3534645
Baumann, J., Hannák, A. and Heitz, C. (2022) ‘Enforcing group fairness in algorithmic decision making : utility maximization under sufficiency’, in Proceedings of the 2022 ACM Conference on Fairness, Accountability, and Transparency. New York: Association for Computing Machinery, pp. 2315–2326. Available at: https://doi.org/10.1145/3531146.3534645.
J. Baumann, A. Hannák, and C. Heitz, “Enforcing group fairness in algorithmic decision making : utility maximization under sufficiency,” in Proceedings of the 2022 ACM Conference on Fairness, Accountability, and Transparency, 2022, pp. 2315–2326. doi: 10.1145/3531146.3534645.
Baumann, Joachim, et al. “Enforcing Group Fairness in Algorithmic Decision Making : Utility Maximization under Sufficiency.” Proceedings of the 2022 ACM Conference on Fairness, Accountability, and Transparency, Association for Computing Machinery, 2022, pp. 2315–26, https://doi.org/10.1145/3531146.3534645.
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.