ABSTRACT
In many real-world situations, data is distributed across multiple self-interested agents. These agents can collaborate to build a machine learning model based on data from multiple agents, potentially reducing the error each experiences. However, sharing models in this way raises questions of fairness: to what extent can the error experienced by one agent be significantly lower than the error experienced by another agent in the same coalition? In this work, we consider two notions of fairness that each may be appropriate in different circumstances: egalitarian fairness (which aims to bound how dissimilar error rates can be) and proportional fairness (which aims to reward players for contributing more data). We similarly consider two common methods of model aggregation, one where a single model is created for all agents (uniform), and one where an individualized model is created for each agent. For egalitarian fairness, we obtain a tight multiplicative bound on how widely error rates can diverge between agents collaborating (which holds for both aggregation methods). For proportional fairness, we show that the individualized aggregation method always gives a small player error that is upper bounded by proportionality. For uniform aggregation, we show that this upper bound is guaranteed for any individually rational coalition (where no player wishes to leave to do local learning).
Supplemental Material
Available for Download
Supplementary material (appendix). Code generates the entire paper in order to define the lemmas that are referenced in the appendix (if this is hard to format, please reach out and I can help!)
- Annie Abay, Yi Zhou, Nathalie Baracaldo, Shashank Rajamoni, Ebube Chuba, and Heiko Ludwig. 2020. Mitigating Bias in Federated Learning. arxiv:2012.02447 [cs.LG]Google Scholar
- Alex Bie, Gautam Kamath, and Vikrant Singhal. 2022. Private Estimation with Public Data. arXiv preprint arXiv:2208.07984 (2022).Google Scholar
- Avrim Blum, Nika Haghtalab, Richard Lanas Phillips, and Han Shao. 2021. One for one, or all for all: Equilibria and optimality of collaboration in federated learning. In International Conference on Machine Learning. PMLR, 1005–1014.Google Scholar
- Franziska Boenisch, Adam Dziedzic, Roei Schuster, Ali Shahin Shamsabadi, Ilia Shumailov, and Nicolas Papernot. 2021. When the Curious Abandon Honesty: Federated Learning Is Not Private. arXiv preprint arXiv:2112.02918 (2021).Google Scholar
- Bhaskar Ray Chaudhury, Linyi Li, Mintong Kang, Bo Li, and Ruta Mehta. 2022. Fairness in Federated Learning via Core-Stability. arXiv preprint arXiv:2211.02091 (2022).Google Scholar
- Yae Jee Cho, Divyansh Jhunjhunwala, Tian Li, Virginia Smith, and Gauri Joshi. 2022. To Federate or Not To Federate: Incentivizing Client Participation in Federated Learning. arXiv preprint arXiv:2205.14840 (2022).Google Scholar
- Rachel Cummings, Hadi Elzayn, Vasilis Gkatzelis, Emmanouil Pountourakis, and Juba Ziani. 2022. Optimal data acquisition with privacy-aware agents. arXiv preprint arXiv:2209.06340 (2022).Google Scholar
- Kate Donahue and Jon Kleinberg. 2021. Model-sharing Games: Analyzing Federated Learning Under Voluntary Participation. AAAI 2021 (2021). https://arxiv.org/abs/2010.00753Google Scholar
- Kate Donahue and Jon Kleinberg. 2021. Optimality and Stability in Federated Learning: A Game-theoretic Approach. In Advances in Neural Information Processing Systems. https://proceedings.neurips.cc/paper/2021/hash/09a5e2a11bea20817477e0b1dfe2cc21-Abstract.htmlGoogle Scholar
- Wei Du, Depeng Xu, Xintao Wu, and Hanghang Tong. 2020. Fairness-aware Agnostic Federated Learning. ArXiv abs/2010.05057 (2020).Google Scholar
- Zhenan Fan, Huang Fang, Zirui Zhou, Jian Pei, Michael P Friedlander, Changxin Liu, and Yong Zhang. 2022. Improving Fairness for Data Valuation in Horizontal Federated Learning. In 2022 IEEE 38th International Conference on Data Engineering (ICDE). IEEE, 2440–2453.Google Scholar
- Felix Grimberg, Mary-Anne Hartley, Sai P Karimireddy, and Martin Jaggi. 2021. Optimal model averaging: Towards personalized collaborative learning. arXiv preprint arXiv:2110.12946 (2021).Google Scholar
- Ernst Hafen, D Kossmann, and A Brand. 2014. Health data cooperatives–citizen empowerment. Methods of information in medicine 53, 02 (2014), 82–86.Google Scholar
- Thomas Hardjono and Alex Pentland. 2019. Data cooperatives: Towards a foundation for decentralized personal data management. arXiv preprint arXiv:1905.08819 (2019).Google Scholar
- Cengis Hasan. 2021. Incentive Mechanism Design for Federated Learning: Hedonic Game Approach. arxiv:2101.09673 [cs.GT]Google Scholar
- Joel Hestness, Sharan Narang, Newsha Ardalani, Gregory Diamos, Heewoo Jun, Hassan Kianinejad, Md Patwary, Mostofa Ali, Yang Yang, and Yanqi Zhou. 2017. Deep learning scaling is predictable, empirically. arXiv preprint arXiv:1712.00409 (2017).Google Scholar
- Peter Kairouz, H. Brendan McMahan, Brendan Avent, Aurélien Bellet, Mehdi Bennis, Arjun Nitin Bhagoji, Keith Bonawitz, Zachary Charles, Graham Cormode, Rachel Cummings, Rafael G. L. D’Oliveira, Salim El Rouayheb, David Evans, Josh Gardner, Zachary Garrett, Adrià Gascón, Badih Ghazi, Phillip B. Gibbons, Marco Gruteser, Zaid Harchaoui, Chaoyang He, Lie He, Zhouyuan Huo, Ben Hutchinson, Justin Hsu, Martin Jaggi, Tara Javidi, Gauri Joshi, Mikhail Khodak, Jakub Konečný, Aleksandra Korolova, Farinaz Koushanfar, Sanmi Koyejo, Tancrède Lepoint, Yang Liu, Prateek Mittal, Mehryar Mohri, Richard Nock, Ayfer Özgür, Rasmus Pagh, Mariana Raykova, Hang Qi, Daniel Ramage, Ramesh Raskar, Dawn Song, Weikang Song, Sebastian U. Stich, Ziteng Sun, Ananda Theertha Suresh, Florian Tramèr, Praneeth Vepakomma, Jianyu Wang, Li Xiong, Zheng Xu, Qiang Yang, Felix X. Yu, Han Yu, and Sen Zhao. 2019. Advances and Open Problems in Federated Learning. arxiv:1912.04977 [cs.LG]Google Scholar
- Samhita Kanaparthy, Manisha Padala, Sankarshan Damle, and Sujit Gujar. 2021. Fair Federated Learning for Heterogeneous Face Data. arXiv preprint arXiv:2109.02351 (2021).Google Scholar
- Justin Kang, Ramtin Pedarsani, and Kannan Ramchandran. 2023. The Fair Value of Data Under Heterogeneous Privacy Constraints. arXiv preprint arXiv:2301.13336 (2023).Google Scholar
- Sai Praneeth Karimireddy, Wenshuo Guo, and Michael I Jordan. 2022. Mechanisms that Incentivize Data Sharing in Federated Learning. arXiv preprint arXiv:2207.04557 (2022).Google Scholar
- Tian Li, Shengyuan Hu, Ahmad Beirami, and Virginia Smith. 2021. Ditto: Fair and robust federated learning through personalization. In International Conference on Machine Learning. PMLR, 6357–6368.Google Scholar
- Tian Li, Anit Kumar Sahu, Ameet Talwalkar, and Virginia Smith. 2020. Federated Learning: Challenges, Methods, and Future Directions. IEEE Signal Processing Magazine 37, 3 (May 2020), 50–60. https://doi.org/10.1109/msp.2020.2975749Google Scholar
- Tian Li, Maziar Sanjabi, Ahmad Beirami, and Virginia Smith. 2019. Fair Resource Allocation in Federated Learning. arxiv:1905.10497 [cs.LG]Google Scholar
- Yiwei Li, Shuai Wang, Chong-Yung Chi, and Tony QS Quek. 2023. Differentially Private Federated Clustering over Non-IID Data. arXiv preprint arXiv:2301.00955 (2023).Google Scholar
- Richard K. Lomotey, Sandra Kumi, and Ralph Deters. 2022. Data Trusts as a Service: Providing a platform for multi‐party data sharing. International Journal of Information Management Data Insights 2, 1 (2022), 100075. https://doi.org/10.1016/j.jjimei.2022.100075Google Scholar
- S. Lozano. 2012. Information sharing in DEA: A cooperative game theory approach. European Journal of Operational Research 222, 3 (2012), 558–565. https://doi.org/10.1016/j.ejor.2012.05.014Google Scholar
- L. Lyu, Xinyi Xu, and Q. Wang. 2020. Collaborative Fairness in Federated Learning. ArXiv abs/2008.12161 (2020).Google Scholar
- H. Brendan McMahan, Eider Moore, Daniel Ramage, Seth Hampson, and Blaise Agüera y Arcas. 2016. Communication-Efficient Learning of Deep Networks from Decentralized Data. arxiv:1602.05629 [cs.LG]Google Scholar
- Mehryar Mohri, Gary Sivek, and Ananda Theertha Suresh. 2019. Agnostic Federated Learning. arxiv:1902.00146 [cs.LG]Google Scholar
- Viraaji Mothukuri, Reza M Parizi, Seyedamin Pouriyeh, Yan Huang, Ali Dehghantanha, and Gautam Srivastava. 2021. A survey on security and privacy of federated learning. Future Generation Computer Systems 115 (2021), 619–640.Google Scholar
- Afroditi Papadaki, Natalia Martinez, Martin Bertran, Guillermo Sapiro, and Miguel Rodrigues. 2021. Federating for Learning Group Fair Models. arXiv preprint arXiv:2110.01999 (2021).Google Scholar
- Yingying Pei, Fen Hou, and Lin X. Cai. 2019. A Stable and Fair Coalition Formation Scheme in Mobile Crowd Sensing. In ICC 2019 - 2019 IEEE International Conference on Communications (ICC). 1–6. https://doi.org/10.1109/ICC.2019.8761948Google Scholar
- Esther Rolf, Theodora T Worledge, Benjamin Recht, and Michael Jordan. 2021. Representation matters: Assessing the importance of subgroup allocations in training data. In International Conference on Machine Learning. PMLR, 9040–9051.Google Scholar
- Yuxin Shi, Han Yu, and Cyril Leung. 2021. A Survey of Fairness-Aware Federated Learning. arXiv preprint arXiv:2111.01872 (2021).Google Scholar
- Tianshu Song, Yongxin Tong, and Shuyue Wei. 2019. Profit Allocation for Federated Learning. 2019 IEEE International Conference on Big Data (Big Data) (2019), 2577–2586.Google Scholar
- Xuezhen Tu, Kun Zhu, Nguyen Cong Luong, Dusit Niyato, Yang Zhang, and Juan Li. 2021. Incentive Mechanisms for Federated Learning: From Economic and Game Theoretic Perspective. arXiv preprint arXiv:2111.11850 (2021).Google Scholar
- Ilse Van Roessel, Matthias Reumann, and Angela Brand. 2017. Potentials and challenges of the health data cooperative model. Public health genomics 20, 6 (2017), 321–331.Google Scholar
- Shanshan Wu, Tian Li, Zachary Charles, Yu Xiao, Ziyu Liu, Zheng Xu, and Virginia Smith. 2022. Motley: Benchmarking heterogeneity and personalization in federated learning. arXiv preprint arXiv:2206.09262 (2022).Google Scholar
- Xinyi Xu and L. Lyu. 2020. Towards Building a Robust and Fair Federated Learning System. ArXiv abs/2011.10464 (2020).Google Scholar
- Yuchen Zeng, Hongxu Chen, and Kangwook Lee. 2021. Improving Fairness via Federated Learning. arXiv preprint arXiv:2110.15545 (2021).Google Scholar
- Fengda Zhang, Kun Kuang, Yuxuan Liu, Chao Wu, Fei Wu, Jiaxun Lu, Yunfeng Shao, and Jun Xiao. 2021. Unified Group Fairness on Federated Learning. arXiv preprint arXiv:2111.04986 (2021).Google Scholar
- Jingfeng Zhang, Cheng Li, A. Robles-Kelly, and M. Kankanhalli. 2020. Hierarchically Fair Federated Learning. ArXiv abs/2004.10386 (2020).Google Scholar
- Jie Zhang, Zhihao Qu, Chenxi Chen, Haozhao Wang, Yufeng Zhan, Baoliu Ye, and Song Guo. 2021. Edge Learning: The Enabling Technology for Distributed Big Data Analytics in the Edge. ACM Comput. Surv. 54, 7, Article 151 (jul 2021), 36 pages. https://doi.org/10.1145/3464419Google Scholar
- Lefeng Zhang, Tianqing Zhu, Ping Xiong, Wanlei Zhou, and Philip S. Yu. 2022. A Game-theoretic Federated Learning Framework for Data Quality Improvement. IEEE Transactions on Knowledge and Data Engineering (2022), 1–15. https://doi.org/10.1109/TKDE.2022.3230959Google Scholar
- Yuming Zhang, Bohao Feng, Wei Quan, Aleteng Tian, Keshav Sood, Youfang Lin, and Hongke Zhang. 2020. Cooperative Edge Caching: A Multi-Agent Deep Learning Based Approach. IEEE Access 8 (2020), 133212–133224. https://doi.org/10.1109/ACCESS.2020.3010329Google Scholar
Index Terms
- Fairness in model-sharing games
Recommendations
Local Fairness in Hedonic Games via Individual Threshold Coalitions
AAMAS '16: Proceedings of the 2016 International Conference on Autonomous Agents & Multiagent SystemsHedonic games are coalition formation games where players only specify preferences over coalitions they are part of. We introduce and systematically study local fairness notions in hedonic games by suitably adapting fairness notions from fair division. ...
Allocating indivisible goods
The problem of allocating divisible goods has enjoyed a lot of attention in both mathematics (e.g. the cake-cutting problem) and economics (e.g. market equilibria). On the other hand, the natural requirement of indivisible goods has been somewhat ...
Fairness In Recurrent Auctions With Competing Markets And Supply Fluctuations
Auctions have been used to deal with resource allocation in multiagent environments, especially in service-oriented electronic markets. In this type of market, resources are perishable and auctions are repeated over time with the same or a very similar ...
Comments