|
|
E-Commerce Letters 2024
面向网络贸易交易的联邦学习最优委托策略分析
|
Abstract:
近年来，数字化经济极大地推动了网络贸易交易的发展。联邦学习技术作为一种分布式的机器学习范式，能够助力网络中分布式的节点进行贸易合作。然而，分布式节点与云服务器之间具备信息不对称性，现有关于联邦学习场景的相关方案存在交互模型构建不准确问题，导致云服务器难以选择最优委托策略。所提出的方法基于信号博弈模型对联邦学习进行建模，并对该场景下的参与者交互问题进行分析。首先，对参与者进行理性假设，定义参与者集合、行动空间、支付函数等博弈要素。其次，引入行为–信念系统，构建服务器与用户之间的信号博弈模型。最后，由序贯理性和序贯一致性，证明该博弈存在序贯均衡，求解服务器针对不同类型用户的最优委托策略。通过实验仿真，验证选择最优委托策略的必要性。
In recent years, the digital economy has greatly promoted the development of online trade transactions. As a distributed machine learning paradigm, federated learning technology can help distributed nodes in the network to conduct trade cooperation. However, there is information asymmetry between distributed nodes and cloud servers. The existing related schemes about federated learning scenarios have the problem of inaccurate interaction model construction, which makes it difficult for cloud servers to choose the optimal delegation strategy. The proposed method models federated learning based on the signal game model, and analyzes the problem of participant interaction in this scenario. Firstly, the rational hypothesis of the participants is carried out, and the game elements such as the set of participants, the action space and the payment function are defined. Secondly, the behavior-belief system is introduced to construct the signal game model between the server and the user. Finally, by sequential rationality and sequential consistency, it is proved that the game has sequential equilibrium, and the optimal delegation strategy of the server for different types of users is solved. Through experimental simulation, the necessity of selecting the optimal delegation strategy is verified.
| [1] | Mammen, P.M. (2021) Federated Learning: Opportunities and Challenges. 1-5. https://doi.org/10.48550/arXiv.2101.05428 |
| [2] | Heiss, J., Grünewald, E., Tai, S., et al. (2022) Advancing Blockchain-Based Federated Learning through Verifiable Off-chain Computations. 2022 IEEE International Conference on Blockchain, Espoo, 22-25 August 2022, 194-201. https://doi.org/10.1109/Blockchain55522.2022.00034 |
| [3] | Data.Kone？ny, J., McMahan, H.B., Ramage, D., et al. (2016) Federated Optimization: Distributed Machine Learning for On-Device Intelligence. 1-38. https://doi.org/10.48550/arXiv.1610.02527 |
| [4] | Banabilah, S., Aloqaily, M., Alsayed, E., et al. (2022) Federated Learning Review: Fundamentals, Enabling Technologies, and Future Applications. Information Processing & Management, 59, Article ID: 103061. https://doi.org/10.1016/j.ipm.2022.103061 |
| [5] | Zhang, K., Song, X., Zhang, C., et al. (2022) Challenges and Future Directions of Secure Federated Learning: A Survey. Frontiers of Computer Science, 16, 1-8. https://doi.org/10.1007/s11704-021-0598-z |
| [6] | Liu, J., Huang, J., Zhou, Y., et al. (2022) From Distributed Machine Learning to Federated Learning: A Survey. Knowledge and Information Systems, 64, 885-917. https://doi.org/10.1007/s10115-022-01664-x |
| [7] | Nguyen, D.C., Pham, Q.V., Pathirana, P.N., et al. (2022) Federated Learning for Smart Healthcare: A Survey. ACM Computing Surveys, 55, 1-37. https://doi.org/10.1145/3501296 |
| [8] | Tang, M. and Wong, V.W.S. (2021) An Incentive Mechanism for Cross-Silo Federated Learning: A Public Goods Perspective. IEEE INFOCOM 2021-IEEE Conference on Computer Communications, Vancouver, BC, 10-13 May 2021, 1-10. https://doi.org/10.1109/INFOCOM42981.2021.9488705 |
| [9] | Pei, J., Zhong, K., Jan, M.A., et al. (2022) Personalized Federated Learning Framework for Network Traffic Anomaly Detection. Computer Networks, 209, Article ID: 108906. https://doi.org/10.1016/j.comnet.2022.108906 |
| [10] | Zhang, T., Gao, L., He, C., et al. (2022) Federated Learning for the Internet of Things: Applications, Challenges, and Opportunities. IEEE Internet of Things Magazine, 5, 24-29. https://doi.org/10.1109/IOTM.004.2100182 |
| [11] | Wang, Z., Hu, Q., Li, R., et al. (2022) Incentive Mechanism Design for Joint Resource Allocation in Blockchain-Based Federated Learning. IEEE Transactions on Parallel and Distributed Systems, 34, 1536-1547. |
| [12] | Nishio, T., Shinkuma, R. and Mandayam, N.B. (2020) Estimation of Individual Device Contributions for Incentivizing Federated Learning. 2020 IEEE Globecom Workshops, Taipei, 7-11 December 2020, 1-6. https://doi.org/10.1109/GCWkshps50303.2020.9367484 |
| [13] | Li, Q., Wen, Z., Wu, Z., et al. (2021) A Survey on Federated Learning Systems: Vision, Hype and Reality for Data Privacy and Protection. IEEE Transactions on Knowledge and Data Engineering, 35, 3347-3366. https://doi.org/10.1109/TKDE.2021.3124599 |
| [14] | Zhang, J., Wu, Y. and Pan, R. (2021) Incentive Mechanism for Horizontal Federated Learning Based on Reputation and Reverse Auction. Proceedings of the Web Conference, Ljubljana, April 2021, 947-956. https://doi.org/10.1145/3442381.3449888 |
