《计算机应用》唯一官方网站 ›› 2024, Vol. 44 ›› Issue (2): 344-352.DOI: 10.11772/j.issn.1001-9081.2023020244
所属专题: 人工智能
收稿日期:
2023-03-09
修回日期:
2023-04-17
接受日期:
2023-04-21
发布日期:
2023-08-14
出版日期:
2024-02-10
通讯作者:
史红周
作者简介:
余孙婕(1998—),女,浙江宁波人,硕士研究生,主要研究方向:信息安全、区块链基金资助:
Sunjie YU1,2, Hui ZENG1,2, Shiyu XIONG1,2, Hongzhou SHI1()
Received:
2023-03-09
Revised:
2023-04-17
Accepted:
2023-04-21
Online:
2023-08-14
Published:
2024-02-10
Contact:
Hongzhou SHI
About author:
YU Sunjie, born in 1998, M. S. candidate. Her research interests include information security, blockchain.Supported by:
摘要:
针对当前联邦学习缺乏公平合理的激励机制,难以衡量不同数据量、不同数据质量、不同数据分布的参与节点的联邦学习贡献度等问题,提出一种基于生成式对抗网络(GAN)的联邦学习激励机制。首先,提出融合训练模型的生成式对抗网络实现高精度样本生成;随后,基于融合训练模型的生成式对抗网络实现激励机制的贡献度评估算法,该算法通过联合模型筛选样本并生成数据标签,引入参与节点的本地数据标签分布平衡非独立同分布数据标签对贡献度评估的影响;最后,使用两阶段Stackelberg博弈实现联邦学习激励过程。安全性分析结果表明,所提激励机制在联邦学习过程中保证数据安全和系统稳定。实验结果表明,所提激励机制具备正确性,贡献度评估算法在不同数据量、不同数据质量和不同数据分布的情况下均有较好的性能。
中图分类号:
余孙婕, 曾辉, 熊诗雨, 史红周. 基于生成式对抗网络的联邦学习激励机制[J]. 计算机应用, 2024, 44(2): 344-352.
Sunjie YU, Hui ZENG, Shiyu XIONG, Hongzhou SHI. Incentive mechanism for federated learning based on generative adversarial network[J]. Journal of Computer Applications, 2024, 44(2): 344-352.
方案 | A1 | A2 | A3 | |||
---|---|---|---|---|---|---|
平均值 | 标准差 | 平均值 | 标准差 | 平均值 | 标准差 | |
文献[ | 62.2 | 16.9 | 5.8 | 2.9 | 1.1 | 0.9 |
文献[ | 100.0 | 0.0 | 99.4 | 0.0 | 0.1 | 0.1 |
文献[ | 4.9 | 3.3 | 84.9 | 1.2 | 99.6 | 0.3 |
文献[ | 48.7 | 24.3 | 41.3 | 14.3 | 33.7 | 11.9 |
文献[ | 88.5 | 6.1 | 49.8 | 3.6 | 8.2 | 4.6 |
本文方案 | 49.2 | 18.9 | 22.9 | 15.8 | 6.7 | 10.2 |
表1 几种方案在不同数据量参与节点时的贡献度统计值
Tab. 1 Comparison of contribution statistics under several scenarios of participate nodes with different data volumes
方案 | A1 | A2 | A3 | |||
---|---|---|---|---|---|---|
平均值 | 标准差 | 平均值 | 标准差 | 平均值 | 标准差 | |
文献[ | 62.2 | 16.9 | 5.8 | 2.9 | 1.1 | 0.9 |
文献[ | 100.0 | 0.0 | 99.4 | 0.0 | 0.1 | 0.1 |
文献[ | 4.9 | 3.3 | 84.9 | 1.2 | 99.6 | 0.3 |
文献[ | 48.7 | 24.3 | 41.3 | 14.3 | 33.7 | 11.9 |
文献[ | 88.5 | 6.1 | 49.8 | 3.6 | 8.2 | 4.6 |
本文方案 | 49.2 | 18.9 | 22.9 | 15.8 | 6.7 | 10.2 |
方案 | B1 | B2 | B3 | |||
---|---|---|---|---|---|---|
平均值 | 标准差 | 平均值 | 标准差 | 平均值 | 标准差 | |
文献[ | 41.8 | 21.2 | 31.1 | 17.9 | 56.4 | 25.0 |
文献[ | 86.3 | 6.4 | 84.8 | 9.3 | 22.7 | 11.2 |
文献[ | 85.8 | 8.0 | 76.5 | 9.0 | 21.3 | 11.5 |
文献[ | 42.6 | 17.9 | 27.5 | 24.9 | 35.0 | 35.7 |
文献[ | 73.2 | 17.5 | 71.8 | 16.7 | 32.5 | 20.8 |
本文方案 | 66.0 | 22.7 | 38.3 | 20.3 | 21.1 | 12.6 |
表2 几种方案在不同数据质量参与节点时的贡献度统计值
Tab. 2 Comparison of contribution statistics under several scenarios of participate nodes with different data qualities
方案 | B1 | B2 | B3 | |||
---|---|---|---|---|---|---|
平均值 | 标准差 | 平均值 | 标准差 | 平均值 | 标准差 | |
文献[ | 41.8 | 21.2 | 31.1 | 17.9 | 56.4 | 25.0 |
文献[ | 86.3 | 6.4 | 84.8 | 9.3 | 22.7 | 11.2 |
文献[ | 85.8 | 8.0 | 76.5 | 9.0 | 21.3 | 11.5 |
文献[ | 42.6 | 17.9 | 27.5 | 24.9 | 35.0 | 35.7 |
文献[ | 73.2 | 17.5 | 71.8 | 16.7 | 32.5 | 20.8 |
本文方案 | 66.0 | 22.7 | 38.3 | 20.3 | 21.1 | 12.6 |
方案 | C1 | C2 | C3 | |||
---|---|---|---|---|---|---|
平均值 | 标准差 | 平均值 | 标准差 | 平均值 | 标准差 | |
文献[ | 59.7 | 20.2 | 15.6 | 15.1 | 59.7 | 24.2 |
文献[ | 92.2 | 4.8 | 69.2 | 6.5 | 56.2 | 23.8 |
文献[ | 51.0 | 10.0 | 21.4 | 10.4 | 49.7 | 23.5 |
文献[ | 45.7 | 17.8 | 48.6 | 26.1 | 42.8 | 27.8 |
文献[ | 89.9 | 8.9 | 67.7 | 10.6 | 18.4 | 10.6 |
本文方案 | 38.8 | 22.7 | 31.1 | 22.8 | 41.4 | 21.4 |
表3 几种方案在不同数据分布参与节点时的贡献度统计值
Tab. 3 Comparison of contribution statistics under several scenarios of participate nodes with different data distributions
方案 | C1 | C2 | C3 | |||
---|---|---|---|---|---|---|
平均值 | 标准差 | 平均值 | 标准差 | 平均值 | 标准差 | |
文献[ | 59.7 | 20.2 | 15.6 | 15.1 | 59.7 | 24.2 |
文献[ | 92.2 | 4.8 | 69.2 | 6.5 | 56.2 | 23.8 |
文献[ | 51.0 | 10.0 | 21.4 | 10.4 | 49.7 | 23.5 |
文献[ | 45.7 | 17.8 | 48.6 | 26.1 | 42.8 | 27.8 |
文献[ | 89.9 | 8.9 | 67.7 | 10.6 | 18.4 | 10.6 |
本文方案 | 38.8 | 22.7 | 31.1 | 22.8 | 41.4 | 21.4 |
样本量 | A1 | A2 | A3 | |||
---|---|---|---|---|---|---|
平均值 | 标准差 | 平均值 | 标准差 | 平均值 | 标准差 | |
120 | 53.6 | 18.5 | 30.4 | 18.1 | 15.6 | 7.2 |
300 | 49.2 | 18.9 | 22.9 | 15.8 | 6.7 | 10.2 |
600 | 49.2 | 22.7 | 19.2 | 17.0 | 6.0 | 10.5 |
表4 不同样本量下参与节点贡献度统计值
Tab. 4 Comparison of contribution statistics of participant nodes with different sample volumes
样本量 | A1 | A2 | A3 | |||
---|---|---|---|---|---|---|
平均值 | 标准差 | 平均值 | 标准差 | 平均值 | 标准差 | |
120 | 53.6 | 18.5 | 30.4 | 18.1 | 15.6 | 7.2 |
300 | 49.2 | 18.9 | 22.9 | 15.8 | 6.7 | 10.2 |
600 | 49.2 | 22.7 | 19.2 | 17.0 | 6.0 | 10.5 |
有无参与节点数据标签分布 | C1 | C2 | C3 | |||
---|---|---|---|---|---|---|
平均值 | 标准差 | 平均值 | 标准差 | 平均值 | 标准差 | |
无 | 54.3 | 29.2 | 39.8 | 29.9 | 31.6 | 26.3 |
有 | 38.8 | 22.7 | 31.1 | 22.8 | 41.4 | 21.4 |
表5 有无参与节点数据标签分布下参与节点贡献度统计值
Tab. 5 Comparison of contribution statistics of participant nodes with or without data label distribution
有无参与节点数据标签分布 | C1 | C2 | C3 | |||
---|---|---|---|---|---|---|
平均值 | 标准差 | 平均值 | 标准差 | 平均值 | 标准差 | |
无 | 54.3 | 29.2 | 39.8 | 29.9 | 31.6 | 26.3 |
有 | 38.8 | 22.7 | 31.1 | 22.8 | 41.4 | 21.4 |
样本生成算法 | B1 | B2 | B3 | |||
---|---|---|---|---|---|---|
平均值 | 标准差 | 平均值 | 标准差 | 平均值 | 标准差 | |
无 | 60.5 | 18.5 | 59.1 | 20.7 | 32.2 | 19.4 |
GAN | 75.5 | 18.0 | 54.4 | 22.8 | 40.5 | 21.5 |
GANT | 66.0 | 22.7 | 38.3 | 20.3 | 21.1 | 12.6 |
表6 不同样本生成下参与节点贡献度统计值
Tab. 6 Comparison of node contribution statistics under different sample generation algorithms
样本生成算法 | B1 | B2 | B3 | |||
---|---|---|---|---|---|---|
平均值 | 标准差 | 平均值 | 标准差 | 平均值 | 标准差 | |
无 | 60.5 | 18.5 | 59.1 | 20.7 | 32.2 | 19.4 |
GAN | 75.5 | 18.0 | 54.4 | 22.8 | 40.5 | 21.5 |
GANT | 66.0 | 22.7 | 38.3 | 20.3 | 21.1 | 12.6 |
1 | 张志刚, 杨栋枢, 吴红侠. 数据资产价值评估模型研究与应用[J]. 现代电子技术, 2015, 38(20): 44-47,51. 10.3969/j.issn.1004-373X.2015.20.012 |
ZHANG Z G, YANG D S, WU H X. Research and application of data asset value assessment model[J]. Modern Electronics Technique, 2015, 38(20): 44-47,51. 10.3969/j.issn.1004-373X.2015.20.012 | |
2 | 张利娟. “数据二十条”出炉,如何挖掘数据“石油”?[J]. 中国报道, 2023(1): 66-68. |
ZHANG L J. "Data Twenty" released, how to mine data "oil"?[J]. China Report, 2023(1): 66-68. | |
3 | McMAHAN H B, MOORE E, RAMAGE D, et al. Communication-efficient learning of deep networks from decentralized data[C]// Proceedings of the 20th International Conference on Artificial Intelligence and Statistics. New York: PMLR, 2017: 1273-1282. |
4 | KUMAR Y, SINGLA R. Federated learning systems for healthcare: Perspective and recent progress[C]// Federated Learning Systems. Studies in Computational Intelligence 965. Cham: Springer, 2021: 141-156. 10.1007/978-3-030-70604-3_6 |
5 | JIANG J C, KANTARCI B, OKTUG S, et al. Federated learning in smart city sensing: challenges and opportunities[J]. Sensors, 2020, 20(21): 6230. 10.3390/s20216230 |
6 | BYRD D, POLYCHRONIADOU A. Differentially private secure multi-party computation for federated learning in financial applications [C]// Proceedings of the 1st ACM International Conference on AI in Finance. New York: ACM, 2020: No. 16. 10.1145/3383455.3422562 |
7 | QU Y, POKHREL S R, GARG S, et al. A blockchained federated learning framework for cognitive computing in industry 4.0 networks[J]. IEEE Transactions on Industrial Informatics, 2020, 17(4): 2964-2973. 10.1109/tii.2020.3007817 |
8 | ZHAN Y, ZHANG J, HONG Z, et al. A survey of incentive mechanism design for federated learning[J]. IEEE Transactions on Emerging Topics in Computing, 2022, 10(2): 1035-1044. |
9 | ZENG R, ZHANG S, WANG J, et al. FMore: an incentive scheme of multi-dimensional auction for federated learning in MEC[C]// Proceedings of the 2020 IEEE 40th International Conference on Distributed Computing Systems. Piscataway: IEEE, 2020: 278-288. 10.1109/icdcs47774.2020.00094 |
10 | SARIKAYA Y, ERCETIN O. Motivating workers in federated learning: a stackelberg game perspective[J]. IEEE Networking Letters, 2019, 2(1): 23-27. 10.1109/lnet.2019.2947144 |
11 | ZHAN Y, LI P, QU Z, et al. A learning-based incentive mechanism for federated learning[J]. IEEE Internet of Things Journal, 2020, 7(7): 6360-6368. 10.1109/jiot.2020.2967772 |
12 | XU X, LYU Y. Towards building a robust and fair federated learning system [EB/OL]. [2023-03-08]. . |
13 | WANG G, DANG C X, ZHOU Z. Measure contribution of participants in federated learning[C]// Proceedings of the 2019 IEEE International Conference on Big Data. Piscataway: IEEE, 2019: 2597-2604. 10.1109/bigdata47090.2019.9006179 |
14 | GOODFELLOW I, POUGET-ABADIE J, MIRZA M, et al. Generative adversarial networks[J]. Communications of the ACM, 2020, 63(11): 139-144. 10.1145/3422622 |
15 | VONDRICK C, PIRSIAVASH H, TORRALBA A. Generating videos with scene dynamics [C]// Proceedings of the 30th International Conference on Neural Information Processing Systems. Red Hook: Curran Associates Inc., 2016: 613-621. 10.1109/cvpr.2016.18 |
16 | REED S, AKATA Z, YAN X, et al. Generative adversarial text to image synthesis [C]// Proceedings of the 33rd International Conference on Machine Learning. New York: JMLR.org, 2016: 1060-1069. |
17 | LI J, MONROE W, SHI T, et al. Adversarial learning for neural dialogue generation [EB/OL]. (2017-09-24)[2023-03-08]. . 10.18653/v1/d17-1230 |
18 | XIN B, YANG W, GENG Y, et al. Private FL-GAN: differential privacy synthetic data generation based on federated learning[C]// Proceedings of the 2020 IEEE International Conference on Acoustics, Speech and Signal Processing. Piscataway: IEEE, 2020: 2927-2931. 10.1109/icassp40776.2020.9054559 |
19 | TABASSUM A, ERBAD A, LEBDA W, et al. FEDGAN-IDS: privacy-preserving IDS using GAN and federated learning[J]. Computer Communications, 2022, 192: 299-310. 10.1016/j.comcom.2022.06.015 |
20 | ZHANG J, ZHANG Q. Stackelberg game for utility-based cooperative cognitive radio networks[C]// Proceedings of the 10th ACM International Symposium on Mobile Ad Hoc Networking and Computing. New York: ACM, 2009: 23-32. 10.1145/1530748.1530753 |
21 | HU R, GONG Y. Trading data for learning: incentive mechanism for on-device federated learning [C]// Proceedings of the 2020 IEEE Global Communications Conference. Piscataway: IEEE, 2020: 1-6. 10.1109/globecom42002.2020.9322475 |
22 | DENG Y, LYU F, REN J, et al. FAIR: quality-aware federated learning with precise user incentive and model aggregation [C]// Proceedings of the 2021 IEEE Conference on Computer Communications. Piscataway: IEEE, 2021: 1-10. 10.1109/infocom42981.2021.9488743 |
23 | HITAJ B, ATENIESE G, PEREZ-CRUZ F. Deep models under the GAN: information leakage from collaborative deep learning [C]// Proceedings of the 2017 ACM SIGSAC conference on computer and communications security. New York: ACM, 2017: 603-618. 10.1145/3133956.3134012 |
24 | KRIZHEVSKY A, HINTON G. Learning multiple layers of features from tiny images[EB/OL]. (2009-04-08)[2023-03-08]. . 10.1016/j.tics.2007.09.004 |
25 | GAO L, LI L, CHEN Y, et al. FIFL: A fair incentive mechanism for federated learning[C]// Proceedings of the 50th International Conference on Parallel Processing. New York: ACM, 2021: No. 82. 10.1145/3472456.3472469 |
26 | LI Y, CHEN C, LIU N, et al. A blockchain-based decentralized federated learning framework with committee consensus[J]. IEEE Network, 2021, 35(1): 234-241. 10.1109/mnet.011.2000263 |
[1] | 陈廷伟, 张嘉诚, 王俊陆. 面向联邦学习的随机验证区块链构建[J]. 《计算机应用》唯一官方网站, 2024, 44(9): 2770-2776. |
[2] | 沈哲远, 杨珂珂, 李京. 基于双流神经网络的个性化联邦学习方法[J]. 《计算机应用》唯一官方网站, 2024, 44(8): 2319-2325. |
[3] | 罗玮, 刘金全, 张铮. 融合秘密分享技术的双重纵向联邦学习框架[J]. 《计算机应用》唯一官方网站, 2024, 44(6): 1872-1879. |
[4] | 陈学斌, 任志强, 张宏扬. 联邦学习中的安全威胁与防御措施综述[J]. 《计算机应用》唯一官方网站, 2024, 44(6): 1663-1672. |
[5] | 张祖篡, 陈学斌, 高瑞, 邹元怀. 基于标签分类的联邦学习客户端选择方法[J]. 《计算机应用》唯一官方网站, 2024, 44(12): 3759-3765. |
[6] | 巫婕, 钱雪忠, 宋威. 基于相似度聚类和正则化的个性化联邦学习[J]. 《计算机应用》唯一官方网站, 2024, 44(11): 3345-3353. |
[7] | 陈学斌, 屈昌盛. 面向联邦学习的后门攻击与防御综述[J]. 《计算机应用》唯一官方网站, 2024, 44(11): 3459-3469. |
[8] | 张帅华, 张淑芬, 周明川, 徐超, 陈学斌. 基于半监督联邦学习的恶意流量检测模型[J]. 《计算机应用》唯一官方网站, 2024, 44(11): 3487-3494. |
[9] | 张欢, 王静宇, 刘立新, 姜晓宇. 双重授权的多组织协同数据共享方案[J]. 《计算机应用》唯一官方网站, 2024, 44(10): 3307-3314. |
[10] | 尹春勇, 周永成. 双端聚类的自动调整聚类联邦学习[J]. 《计算机应用》唯一官方网站, 2024, 44(10): 3011-3020. |
[11] | 张佩瑶, 付晓东. 防恶意竞价的众包多任务分配激励机制[J]. 《计算机应用》唯一官方网站, 2024, 44(1): 261-268. |
[12] | 周辉, 陈玉玲, 王学伟, 张洋文, 何建江. 基于生成对抗网络的联邦学习深度影子防御方案[J]. 《计算机应用》唯一官方网站, 2024, 44(1): 223-232. |
[13] | 蓝梦婕, 蔡剑平, 孙岚. 非独立同分布数据下的自正则化联邦学习优化方法[J]. 《计算机应用》唯一官方网站, 2023, 43(7): 2073-2081. |
[14] | 潘亚楠, 潘庆先, 于兆一, 褚佳静, 于嵩. 时空众包中基于质量感知的在线激励机制[J]. 《计算机应用》唯一官方网站, 2023, 43(7): 2091-2099. |
[15] | 陈宛桢, 张恩, 秦磊勇, 洪双喜. 边缘计算下基于区块链的隐私保护联邦学习算法[J]. 《计算机应用》唯一官方网站, 2023, 43(7): 2209-2216. |
阅读次数 | ||||||
全文 |
|
|||||
摘要 |
|
|||||