QUANTIFYING ALGORITHMIC FAIRNESS: A NOVEL PERSPECTIVE THROUGH UNCERTAINTY ESTIMATION

• Dr. Clara Moreau

  Department of Computer Science, École Normale Supérieure, France

  Author

• Prof. Julia Weber

  Department of Computer Science, École Normale Supérieure, France

  Author

The increasing deployment of machine learning (ML) systems in high-stakes domains necessitates robust fairness evaluation. Traditional fairness metrics primarily focus on statistical disparities in outcomes, often overlooking the model's confidence in its predictions, particularly for sensitive subgroups. This article proposes a novel framework for assessing algorithmic fairness by integrating uncertainty quantification (UQ) into the evaluation process. We delineate between aleatoric (data-inherent) and epistemic (model-inherent) uncertainties and explore various UQ techniques, including Bayesian Neural Networks, Monte Carlo Dropout, Deep Ensembles, and Deep Deterministic Uncertainty. We argue that disparate levels of uncertainty across demographic groups can serve as a powerful diagnostic tool, indicating issues such as data scarcity, representational bias, or inherent ambiguities within specific populations. By leveraging uncertainty as a fairness measure, we can identify subtle forms of discrimination, enhance model transparency, and enable more proactive and targeted bias mitigation strategies. This approach promises to yield more robust, trustworthy, and equitable ML systems.

[1] Abdar, M., Pourpanah, F., Hussain, S., Rezazadegan, D., Liu, L., Ghavamzadeh, M., Fieguth, P., Cao, X., Khosravi, A., Acharya, U. R., et al. (2021). A review of uncertainty quantification in deep learning: Techniques, applications and challenges. Information fusion, 76, 243–297.

[2] Angwin, J., Larson, J., Mattu, S., & Kirchner, L. (2022). Machine bias. In Ethics of data and analytics, pp. 254–264. Auerbach Publications.

[3] Baltaci, Z. S., Oksuz, K., Kuzucu, S., Tezoren, K., Konar, B. K., Ozkan, A., Akbas, E., & Kalkan, S. (2023). Class uncertainty: A measure to mitigate class imbalance. In arXiv preprint arXiv:2311.14090.

[4] Barocas, S., Hardt, M., & Narayanan, A. (2017). Fairness in machine learning. NeurIPS Tutorial, 1, 2.

[5] Becker, B., & Kohavi, R. (1996). Adult. UCI Machine Learning Repository. DOI: https://doi.org/10.24432/C5XW20.

[6] Blundell, C., Cornebise, J., Kavukcuoglu, K., & Wierstra, D. (2015). Weight uncertainty in neural networks. In International conference on machine learning, pp. 1613–1622. PMLR.

[7] Buolamwini, J., & Gebru, T. (2018). Gender shades: Intersectional accuracy disparities in commercial gender classification. In Conference on fairness, accountability and transparency, pp. 77–91. PMLR.

[8] Castelnovo, A., Crupi, R., Greco, G., Regoli, D., Penco, I. G., & Cosentini, A. C. (2022). A clarification of the nuances in the fairness metrics landscape. Scientific Reports, 12(1), 4209.

[9] Cetinkaya, B., Kalkan, S., & Akbas, E. (2024). Ranked: Addressing imbalance and uncertainty in edge detection using ranking-based losses. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp. 3239–3249.

[10] Chen, Y., Raab, R., Wang, J., & Liu, Y. (2022). Fairness transferability subject to bounded distribution shift. Advances in Neural Information Processing Systems, 35, 11266–11278.

[11] Chen, Y., & Joo, J. (2021). Understanding and mitigating annotation bias in facial expression recognition. In Proceedings of the IEEE/CVF International Conference on Computer Vision, pp. 14980–14991.

[12] Cheong, J., Kalkan, S., & Gunes, H. (2021). The hitchhiker’s guide to bias and fairness in facial affective signal processing: Overview and techniques. IEEE Signal Processing Magazine, 38(6), 39–49.

[13] Cheong, J., Kalkan, S., & Gunes, H. (2022). Counterfactual fairness for facial expression recognition. In European Conference on Computer Vision, pp. 245–261. Springer.

[14] Cheong, J., Kalkan, S., & Gunes, H. (2023). Causal structure learning of bias for fair affect recognition. In Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision, pp. 340–349.

[15] Cheong, J., Kalkan, S., & Gunes, H. (2024). Fairrefuse: Referee-guided fusion for multi-modal causal fairness in depression detection. In International Joint Conference on Artificial Intelligence (IJCAI).

[16] Cheong, J., Kuzucu, S., Kalkan, S., & Gunes, H. (2023). Towards gender fairness for mental health prediction. In 32nd Int. Joint Conf. on Artificial Intelligence (IJCAI).

[17] Chouldechova, A. (2017). Fair prediction with disparate impact: A study of bias in recidivism prediction instruments. Big data, 5(2), 153–163.

[18] Ding, F., Hardt, M., Miller, J., & Schmidt, L. (2021). Retiring adult: New datasets for fair machine learning. Advances in neural information processing systems, 34, 6478–6490.

[19] Domnich, A., & Anbarjafari, G. (2021). Responsible ai: Gender bias assessment in emotion recognition..

[20] Dressel, J., & Farid, H. (2018). The accuracy, fairness, and limits of predicting recidivism. Science advances, 4(1), eaao5580.

[21] Dwork, C., Hardt, M., Pitassi, T., Reingold, O., & Zemel, R. (2012). Fairness through awareness. In Proceedings of the 3rd innovations in theoretical computer science conference, pp. 214–226.

[22] Ethayarajh, K. (2020). Is your classifier actually biased? measuring fairness under uncertainty with bernstein bounds. In Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics, pp. 2914–2919.

[23] Feldman, M., Friedler, S. A., Moeller, J., Scheidegger, C., & Venkatasubramanian, S. (2015). Certifying and removing disparate impact. In proceedings of the 21th ACM SIGKDD international conference on knowledge discovery and data mining, pp. 259–268.

[24] Gal, Y., et al. (2016). Uncertainty in deep learning. Ph.D. thesis, University of Cambridge.

[25] Gal, Y., & Ghahramani, Z. (2016). Dropout as a bayesian approximation: Representing model uncertainty in deep learning..

[26] Garg, P., Villasenor, J., & Foggo, V. (2020). Fairness metrics: A comparative analysis. In IEEE International Conference on Big Data (Big Data), pp. 3662–3666. IEEE.

[27] Gawlikowski, J., Tassi, C. R. N., Ali, M., Lee, J., Humt, M., Feng, J., Kruspe, A., Triebel, R., Jung, P., Roscher, R., et al. (2021). A survey of uncertainty in deep neural networks..

[28] Goel, N., Amayuelas, A., Deshpande, A., & Sharma, A. (2021). The importance of modeling data missingness in algorithmic fairness: A causal perspective. In Proceedings of the AAAI Conference on Artificial Intelligence, pp. 7564–7573.

[29] Guo, C., Pleiss, G., Sun, Y., & Weinberger, K. Q. (2017). On calibration of modern neural networks. In International conference on machine learning, pp. 1321–1330. PMLR.

[30] Han, M., Canli, I., Shah, J., Zhang, X., Dino, I. G., & Kalkan, S. (2024). Perspectives of machine learning and natural language processing on characterizing positive energy districts. Buildings, 14(2), 371.

[31] Hardt, M., Price, E., & Srebro, N. (2016). Equality of opportunity in supervised learning. Advances in neural information processing systems, 29.

[32] Havasi, M., Jenatton, R., Fort, S., Liu, J. Z., Snoek, J., Lakshminarayanan, B., Dai, A. M., & Tran, D. (2020). Training independent subnetworks for robust prediction. In International Conference on Learning Representations.

[33] Hort, M., Chen, Z., Zhang, J. M., Harman, M., & Sarro, F. (2023). Bias mitigation for machine learning classifiers: A comprehensive survey. In ACM J. Responsib. Comput., New York, NY, USA. Association for Computing Machinery.

[34] Jiang, H., & Nachum, O. (2020). Identifying and correcting label bias in machine learning. In International Conference on Artificial Intelligence and Statistics, pp. 702–712. PMLR.

[35] Kaiser, P., Kern, C., & R ̈ugamer, D. (2022). Uncertainty-aware predictive modeling for fair data-driven decisions..

[36] Kang, M., Li, L., Weber, M., Liu, Y., Zhang, C., & Li, B. (2022). Certifying some distributional fairness with subpopulation decomposition. Advances in Neural Information Processing Systems, 35, 31045–31058.

[37] Kearns, M., Neel, S., Roth, A., & Wu, Z. S. (2018). Preventing fairness gerrymandering: Auditing and learning for subgroup fairness. In International conference on machine learning, pp. 2564–2572. PMLR.

[38] Kendall, A., & Gal, Y. (2017). What uncertainties do we need in bayesian deep learning for computer vision?. CoRR, abs/1703.04977.

[39] Kingma, D. P., & Ba, J. (2017). Adam: A method for stochastic optimization..

[40] Kullback, S., & Leibler, R. A. (1951). On information and sufficiency. The annals of mathematical statistics, 22(1), 79–86.

[41] Kusner, M. J., Loftus, J., Russell, C., & Silva, R. (2017). Counterfactual fairness. Advances in neural information processing systems, 30.

[42] Kwon, Y., Won, J.-H., Kim, B. J., & Paik, M. C. (2020). Uncertainty quantification using bayesian neural networks in classification: Application to biomedical image segmentation. Computational Statistics & Data Analysis, 142, 106816.

[43] Lakshminarayanan, B., Pritzel, A., & Blundell, C. (2017). Simple and scalable predictive uncertainty estimation using deep ensembles..

[44] Liu, J. Z., Lin, Z., Padhy, S., Tran, D., Bedrax-Weiss, T., & Lakshminarayanan, B. (2020). Simple and principled uncertainty estimation with deterministic deep learning via distance awareness. CoRR, abs/2006.10108.

[45] MacKay, D. J. (1992). A practical bayesian framework for backpropagation networks. Neural computation, 4(3), 448–472.

[46] Mehrabi, N., Morstatter, F., Saxena, N., Lerman, K., & Galstyan, A. (2021). A survey on bias and fairness in machine learning. ACM computing surveys (CSUR), 54(6), 1–35.

[47] Mehta, R., Shui, C., & Arbel, T. (2023). Evaluating the fairness of deep learning uncertainty estimates in medical image analysis..

[48] Mukherjee, D., Yurochkin, M., Banerjee, M., & Sun, Y. (2020). Two simple ways to learn individual fairness metrics from data. In International Conference on Machine Learning, pp. 7097–7107. PMLR.

[49] Mukhoti, J., Kirsch, A., van Amersfoort, J., Torr, P. H., & Gal, Y. (2023). Deep deterministic uncertainty: A new simple baseline. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pp. 24384–24394.

[50] Mukhoti, J., Kulharia, V., Sanyal, A., Golodetz, S., Torr, P., & Dokania, P. (2020). Calibrating deep neural networks using focal loss. In Larochelle, H., Ranzato, M., Hadsell, R., Balcan, M., & Lin, H. (Eds.), Advances in Neural Information Processing Systems, Vol. 33, pp. 15288–15299. Curran Associates, Inc.

[51] Naik, L., Kalkan, S., & Kruger, N. (2024). Pre-grasp approaching on mobile robots: A pre-active layered approach. IEEE Robotics and Automation Letters, 9(3).

[52] Neal, R. M. (1995). Bayesian Learning for Neural Networks. Ph.D. thesis, University of Toronto.

[53] Roy, A., & Mohapatra, P. (2023). Fairness uncertainty quantification: How certain are you that the model is fair?..

[54] Rudin, C. (2019). Stop explaining black box machine learning models for high stakes decisions and use interpretable models instead. Nature machine intelligence, 1(5), 206–215.

[55] Shridhar, K., Laumann, F., & Liwicki, M. (2019). A comprehensive guide to bayesian convolutional neural network with variational inference. CoRR, abs/1901.02731.

[56] Tahir, A., Cheng, L., & Liu, H. (2023). Fairness through aleatoric uncertainty..

[57] van Amersfoort, J., Smith, L., Teh, Y. W., & Gal, Y. (2020a). Simple and scalable epistemic uncertainty estimation using a single deep deterministic neural network. CoRR, abs/2003.02037.

[58] Van Amersfoort, J., Smith, L., Teh, Y. W., & Gal, Y. (2020b). Uncertainty estimation using a single deep deterministic neural network. In International conference on machine learning, pp. 9690–9700. PMLR.

[59] Verma, S., & Rubin, J. (2018a). Fairness definitions explained. In Proceedings of the international workshop on software fairness, pp. 1–7.

[60] Verma, S., & Rubin, J. (2018b). Fairness definitions explained. In Proceedings of the International Workshop on Software Fairness, FairWare ’18, p. 1–7, New York, NY, USA. Association for Computing Machinery.

[61] Wang, H., He, L., Gao, R., & Calmon, F. (2023). Aleatoric and epistemic discrimination: Fundamental limits of fairness interventions. In Thirty-seventh Conference on Neural Information Processing Systems.

[62] Wang, J., Liu, Y., & Levy, C. (2021). Fair classification with group-dependent label noise. In Proceedings of the 2021 ACM conference on fairness, accountability, and transparency, pp. 526–536.

[63] Xu, T., White, J., Kalkan, S., & Gunes, H. (2020). Investigating bias and fairness in facial expression recognition. In Computer Vision–ECCV 2020 Workshops: Glasgow, UK, August 23–28, 2020, Proceedings, Part VI 16, pp. 506–523. Springer.

[64] Yoon, J., Kang, C., Kim, S., & Han, J. (2022). D-vlog: Multimodal vlog dataset for depression detection. Proceedings of the AAAI Conference on Artificial Intelligence, 36(11), 12226–12234.

[65] Zafar, M. B., Valera, I., Gomez Rodriguez, M., & Gummadi, K. P. (2017). Fairness beyond disparate treatment & disparate impact: Learning classification without disparate mistreatment. In Proceedings of the 26th international conference on world wide web, pp. 1171–1180.

[66] Zanna, K., Sridhar, K., Yu, H., & Sano, A. (2022). Bias reducing multitask learning on mental health prediction..

[67] Zemel, R., Wu, Y., Swersky, K., Pitassi, T., & Dwork, C. (2013). Learning fair representations. In International conference on machine learning, pp. 325–333. PMLR.

• pdf

All articles published by The Parthenon Frontiers and its associated journals are distributed under the terms of the Creative Commons Attribution (CC BY 4.0) International License unless otherwise stated. 

Authors retain full copyright of their published work. By submitting their manuscript, authors agree to grant The Parthenon Frontiers a non-exclusive license to publish, archive, and distribute the article worldwide. Authors are free to:

• Share their article on personal websites, institutional repositories, or social media platforms.

• Reuse their content in future works, presentations, or educational materials, provided proper citation of the original publication.