Authors : Vishal Yadav; Shreeja Kale

Volume/Issue : Volume 9 - 2024, Issue 8 - August

Google Scholar : https://tinyurl.com/28kjr4ak

Scribd : https://tinyurl.com/yvx2epyk

DOI : https://doi.org/10.38124/ijisrt/IJISRT24AUG1319

Abstract : Federated Learning (FL) enables collaborative model training across decentralized devices while preserving user data privacy. However, disparities in data distributions among clients can lead to biased models that perform unfairly across different demographic groups. This paper proposes a fairness-aware Federated Learning framework equipped with real-time bias detection and correction mechanisms. Our approach adjusts model updates to address biases detected at local client levels before aggregating them at the central server. We demonstrate the effectiveness of our method through empirical evaluations on multiple datasets, showcasing significant improvements in fairness and model accuracy. Our proposed framework involves a multi-tiered approach to ensure fairness in the model training process. Firstly, it employs local bias detection techniques at the client level to identify disparities in model performance across different groups. Clients then utilize bias correction mechanisms to adjust their model updates, addressing any detected biases before sending updates to the central server. The central server aggregates these bias-corrected updates, ensuring that the global model benefits from equitable learning while maintaining overall performance.

Keywords : Bias Detection, Real-Time Systems, Fairness- Aware Learning.

References :

1. Bonawitz, K., Eichner, H., Grieskamp, W., et al. (2019). "Towards Federated Learning at Scale: System Design." Proceedings of the 2nd SysML Conference.
2. Hard, A., Rao, K., Mathews, R., Ramaswamy, S. (2018). "Federated Learning for Mobile Keyboard Prediction." arXiv preprint arXiv:1811.03604.
3. Dwork, C., Hardt, M., Pitassi, T., Reingold, O., Zemel, R. (2012). "Fairness through Awareness." Proceedings of the 3rd Innovations in Theoretical Computer Science Conference.
4. Zafar, M. B., Valera, I., Rodriguez, M. G., & Gummadi, K. P. (2017). "Fairness Constraints: Mechanisms for Fair Classification." Proceedings of the 20th International Conference on Artificial Intelligence and Statistics (AISTATS).
5. Agarwal, A., Dudik, M., & Wu, Z. S. (2018). "Fair Regression: Quantitative Definitions and Reduction-Based Algorithms." Proceedings of the 35th International Conference on Machine Learning (ICML).
6. Y McMahan, H. B., Moore, E., Ramage, D., Hampson, S., & Arcas, B. A. y. (2017). "Communication-Efficient Learning of Deep Networks from Decentralized Data." Proceedings of the 20th International Conference on Artificial Intelligence and Statistics (AISTATS).
7. Zemel, R., Wu, Y., Swersky, K., Pitassi, T., & Dwork, C. (2013). "Learning Fair Representations." Proceedings of the 30th International Conference on Machine Learning (ICML).