FDML: A Collaborative Machine Learning Framework for Distributed Features
The DOI matches its Crossref record; 24 of the 26 references it cites were found in a registry or an index.
Abstract
Most current distributed machine learning systems try to scale up model training by using a data-parallel architecture that divides the computation for different samples among workers. We study distributed machine learning from a different motivation, where the information about the same samples, e.g., users and objects, are owned by several parities that wish to collaborate but do not want to share raw data with each other. We propose an asynchronous stochastic gradient descent (SGD) algorithm for such a feature distributed machine learning (FDML) problem, to jointly learn from distributed features, with theoretical convergence guarantees under bounded asynchrony. Our algorithm does not require sharing the original features or even local model parameters between parties, thus preserving the data locality. The system can also easily incorporate differential privacy mechanisms to preserve a higher level of privacy. We implement the FDML system in a parameter server architecture and compare our system with fully centralized learning (which violates data locality) and learning based on only local features, through extensive experiments performed on both a public data set a9a, and a large dataset of 5,000,000 records and 8700 decentralized features from three collaborating apps at Tencent including Tencent MyApp, Tecent QQ Browser and Tencent Mobile Safeguard. Experimental results have demonstrated that the proposed FDML system can be used to significantly enhance app recommendat…
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
In the citing paper
“Therefore, the data features and the associated model parametric blocks are split among edge devices, for which the asynchronous SGD method can be applied for vertical FL [157].”
From Edge artificial intelligence for 6g: vision, enabling technologies, and applications (2022) · cites this paper as [157]
Citations
Cites
Communication-Efficient Learning of Deep Networks from Decentralized Data
McMahan, H. B., Moore, E., Ramage, D., Hampson, S., & Arcas, B. A. Y. (2016) · arXiv · doi:10.48550/arxiv.1602.05629 · Cited by 14,215 · cited 1×
“As a result, a number of distributed machine learning techniques have been proposed to collaboratively train a model by letting each party perform local model updates and exchange locally computed gradients [27] or model parameters [21] with the central server to iteratively improve model accuracy.”
BackgroundThe Algorithmic Foundations of Differential Privacy
Dwork, C., & Roth, A. (2014) · Foundations and Trends® in Theoretical Computer Science · doi:10.1561/0400000042 · Cited by 5,557 · cited 2×
“Moreover, we further enhance the data privacy by adopting a differential-privacy-based method [9, 10, 27].”
MethodLarge Scale Distributed Deep Networks
Dean, J. B., Corrado, G. S., Monga, R., Chen, K., Devin, M., Mao, M., Ranzato, M., Senior, A., Tucker, P. A., Yang, K., et al. (2012) · neural information processing systems · Cited by 4,489
The Algorithmic Foundations of Differential Privacy
Dwork, C., & Roth, A. (2013) · doi:10.1561/9781601988195 · Cited by 3,856
Differential Privacy: A Survey of Results
Dwork, C. (2008) · Lecture Notes in Computer Science · doi:10.1007/978-3-540-79228-4_1 · Cited by 3,542 · cited 2×
“Moreover, we further enhance the data privacy by adopting a differential-privacy-based method [9, 10, 27].”
MethodPrivacy-preserving deep learning
Shokri, R., & Shmatikov, V. (2015) · Proceedings of the 22nd ACM SIGSAC Conference on Computer and Communications Security · doi:10.1145/2810103.2813687 · Cited by 2,356 · cited 6×
“To further protect the feature data at each party from malicious servers and parties, we apply differential privacy based methods by perturbing the local predictions to be uploaded [9, 10, 27].”
BackgroundMethodInfluentialSecureML: A System for Scalable Privacy-Preserving Machine Learning
Mohassel, P., & Zhang, Y. (2017) · 2017 IEEE Symposium on Security and Privacy (SP) · doi:10.1109/sp.2017.12 · Cited by 1,936 · cited 1×
“, via homomorphic encryption [11, 28] or secret sharing [3, 22, 29].”
BackgroundDeep Models Under the GAN: Information Leakage from Collaborative Deep Learning
Hitaj, B., Ateniese, G., & Perez-Cruz, F. (2017) · Proceedings of the 2017 ACM SIGSAC Conference on Computer and Communications Security · doi:10.1145/3133956.3134012 · Cited by 1,377 · cited 2×
“Although raw data are not transferred by the distributed SGD in [27], a recent study [12] points out that an algorithm that passes model parameters may be vulnerable to model inversion attacks based on generative adversarial networks (GANs).”
Background(2016) · International Conference on Machine Learning · Cited by 1,270
Hogwild: A Lock-Free Approach to Parallelizing Stochastic Gradient Descent
Niu, F., Recht, B., Re, C., & Wright, S. J. (2011) · arXiv · doi:10.48550/arxiv.1106.5730 · Cited by 1,211 · cited 1×
“The convergence results have been developed for both gradient descent and SGD [13, 20, 25] as well as proximal gradient methods [18] under different assumptions of the loss functions.”
Method
Research works printed but not linked: 14
- Abadi Mart'in, 2016, Proc. USENIX Symposium on Operating System Design and Implementation (OSDI) .
- Ben-Nun Tal, 2018, Demystifying Parallel and Distributed Deep Learning: An In-Depth Concurrency Analysis. arXiv preprint arXiv:1802.09941 · arXiv
- Bradley Joseph K, 2011, Parallel coordinate descent for l1-regularized loss minimization. arXiv preprint arXiv:1105.5379 · arXiv
- Dean Jeffrey, 2012, et almbox
- 10.5555/2503308.2188391 · doi:10.5555/2503308.2188391
- 10.5555/1791834.1791836 · doi:10.5555/1791834.1791836
- Gilad-Bachrach Ran, 2016, International Conference on Machine Learning . 201--210
- Ho Qirong, 2013, Phillip B Gibbons, Garth A Gibson, Greg Ganger, and Eric P Xing.
- Hsieh Kevin, 2017, Gaia: Geo-Distributed Machine Learning Approaching LAN Speeds.. In NSDI. 629--647.
- Lee Seunghak, 2014, Xun Zheng, Qirong Ho, Garth A Gibson, and Eric P Xing.
- Xiangru Lian Yijun Huang Yuncheng Li and Ji Liu. 2015. Asynchronous parallel stochastic gradient for nonconvex optimization. In Advances in Neural Information Processing Systems. 2737--2745. Xiangru Lian Yijun Huang Yuncheng Li and Ji Liu. 2015. Asynchronous parallel stochastic gradient for nonconvex optimization. In Advances in Neural Information Processing Systems. 2737--2745.
- McMahan H Brendan, 2016, et almbox
- Chad Scherrer Ambuj Tewari Mahantesh Halappanavar and David Haglin. 2012. Feature clustering for accelerating parallel coordinate descent. In Advances in Neural Information Processing Systems. 28--36. Chad Scherrer Ambuj Tewari Mahantesh Halappanavar and David Haglin. 2012. Feature clustering for accelerating parallel coordinate descent. In Advances in Neural Information Processing Systems. 28--36.
- Takabi Hassan, 2016, 29th Annual Conference on Neural Information Processing Systems (NIPS) .
Cited by
A Survey on Federated Learning Systems: Vision, Hype and Reality for Data Privacy and Protection
Li, Q., Wen, Z., Wu, Z., Hu, S., Wang, N., Li, Y., Liu, X., & He, B. (2023) · IEEE Transactions on Knowledge and Data Engineering · doi:10.1109/tkde.2021.3124599 · Cited by 1,407
“[71] propose FDML for vertical FL assuming all parties have the labels.”
BackgroundEdge artificial intelligence for 6g: vision, enabling technologies, and applications
Letaief, K. B., Shi, Y., Lu, J., & Lu, J. (2022) · IEEE Journal on Selected Areas in Communications · doi:10.1109/jsac.2021.3126076 · Cited by 829
“Therefore, the data features and the associated model parametric blocks are split among edge devices, for which the asynchronous SGD method can be applied for vertical FL [157].”
MethodVertical Federated Learning: Concepts, Advances, and Challenges
Liu, Y., Kang, Y., Zou, T., Pu, Y., He, Y., Ye, X., Ouyang, Y., Zhang, Y.-Q., & Yang, Q. (2024) · IEEE Transactions on Knowledge and Data Engineering · doi:10.1109/tkde.2024.3352628 · Cited by 378
“[27] proposed FDML, allowing each party to update its local model asynchronously but based on the same sequence of randomly sampled training data.”
BackgroundMethodInfluentialA Survey of Incentive Mechanism Design for Federated Learning
Zhan, Y., Zhang, J., Hong, Z., Wu, L., Li, P., & Guo, S. (2021) · IEEE Transactions on Emerging Topics in Computing · doi:10.1109/tetc.2021.3063517 · Cited by 284
FedBCD: A Communication-Efficient Collaborative Learning Framework for Distributed Features
Liu, Y., Zhang, X., Kang, Y., Li, L., Chen, T., Hong, M., & Yang, Q. (2022) · IEEE Transactions on Signal Processing · doi:10.1109/tsp.2022.3198176 · Cited by 308
“Feature-partitionedcollaborativelearningproblemshave been studied in the setting of both DL [4], [5], [6] and FL [7], [8], [9], [10].”
BackgroundFeature Inference Attack on Model Predictions in Vertical Federated Learning
Luo, X., Wu, Y., Xiao, X., & Ooi, B. C. (2021) · 2021 IEEE 37th International Conference on Data Engineering (ICDE) · doi:10.1109/icde51399.2021.00023 · Cited by 212
“With horizontal FL being thoroughly studied [7], vertical FL is receiving increased attention recently [4, 8, 10, 11, 26, 37].”
BackgroundVAFL: a Method of Vertical Asynchronous Federated LearningPreprint
Chen, T., Jin, X. P., Sun, Y., & Yin, W. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2007.06081 · Cited by 216
“Compared to the relatively well-studied horizontal FL setting [25], the vertical FL setting has its unique features and challenges [15; 18].”
BackgroundFederated Recommendation Systems
Yang, L., Tan, B., Zheng, V. W., Chen, K., & Yang, Q. (2020) · Lecture Notes in Computer Science · doi:10.1007/978-3-030-63076-8_16 · Cited by 191
“[10] proposed an asynchronous stochastic gradient descent algorithm.”
BackgroundPrivacy preserving vertical federated learning for tree-based models
Wu, Y., Cai, S., Xiao, X., Chen, G., & Ooi, B. C. (2020) · Proceedings of the VLDB Endowment · doi:10.14778/3407790.3407811 · Cited by 184
“In particular, [40, 65] assume that the labels in the training data could be shared with all participating clients in plaintext, whereas in practice, the labels often exist in one client’s data only and could not be revealed to other clients without violating privacy.”
BackgroundGu, B., Xu, A., Huo, Z., Deng, C., & Huang, H. (2022) · IEEE Transactions on Neural Networks and Learning Systems · doi:10.1109/tnnls.2021.3072238 · Cited by 163
“This has also been shown in [34] with the logistic regression model.”
MethodInfluentialDuan, Q., Hu, S., Deng, R., & Lu, Z. (2022) · Sensors · doi:10.3390/s22165983 · Cited by 108
“…Average (FedAvg) algorithm (given in Algorithm 1) [9] is the most widely accepted algorithm for basic in vertical FL are referred to [10], [11], [12], and [13] for each client k ∈ S t in parallel do number of devices each of which contributes a relatively small set of data for training ML…”
Federated Learning Meets Natural Language Processing: A SurveyPreprint
Liu, M., Ho, S., Wang, M., Gao, L., Jin, Y., & Zhang, H. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2107.12603 · Cited by 129
“[26] proposed an asynchronous stochastic gradient descent algorithm.”
Background${\sf FederBoost}$: Private Federated Learning for GBDT
Tian, Z., Zhang, R., Hou, X., Lyu, L., Zhang, T., Liu, J., & Ren, K. (2023) · IEEE Transactions on Dependable and Secure Computing · doi:10.1109/tdsc.2023.3276365 · Cited by 111
“In another work named FDML [16], each participant trains a separate model using its own feature space and shares the prediction result to others (through a central server).”
Salau, B. A., Rawal, A., & Rawat, D. B. (2022) · IEEE Internet of Things Journal · doi:10.1109/jiot.2022.3170449 · Cited by 106
“Convergence solution have been proposed for both gradient descent and Stochastic gradient descent [40], [41].”
Distributed Learning in the Nonconvex World: From batch data to streaming and beyond
Chang, T., Hong, M., Wai, H.-T., Zhang, X., & Lu, S. (2020) · IEEE Signal Processing Magazine · doi:10.1109/msp.2020.2970170 · Cited by 114
“This imposes the challenging question of how to enable the agents that own different sets of features to collaborate on the learning task; see [16] and [38].”
BackgroundPrivacy Intelligence: A Survey on Image Privacy in Online Social Networks
Liu, C., Zhu, T., Zhang, J., & Zhou, W. (2022) · ACM Computing Surveys · doi:10.1145/3547299 · Cited by 102
“The distributed learning frameworks such as collaborative learning [242]–[244] and federated learning [245] may be candidates, as they develop independent models locally without data sharing, then aggregate the local models together to attain a global model.”
BackgroundA federated data-driven evolutionary algorithm
Xu, J., Jin, Y., Du, W., & Gu, S. (2021) · Knowledge-Based Systems · doi:10.1007/s40747-021-00506-7 · Cited by 80
VF 2 Boost: Very Fast Vertical Federated Gradient Boosting for Cross-Enterprise Learning
Fu, F., Shao, Y., Yu, L., Jiang, J., Xue, H., Tao, Y., & Cui, B. (2021) · ACM SIGMOD International Conference on Management of Data (SIGMOD) · doi:10.1145/3448016.3457241 · Cited by 117
“Many works are also raised for the case where two or more parties expect to jointly train on vertically partitioned data [15, 37, 52, 75, 78, 80, 84, 90].”
BackgroundSecure Bilevel Asynchronous Vertical Federated Learning with Backward Updating
Zhang, Q., Gu, B., Deng, C. D., & Huang, H. (2021) · Proceedings of the AAAI Conference on Artificial Intelligence · doi:10.1609/aaai.v35i12.17301 · Cited by 95
Defending Batch-Level Label Inference and Replacement Attacks in Vertical Federated Learning
Zou, T., Liu, Y., Kang, Y., Liu, W., He, Y., Yi, Z., Yang, Q., & Zhang, Y.-Q. (2022) · IEEE Transactions on Big Data · doi:10.1109/tbdata.2022.3192121 · Cited by 113
BlindFL: Vertical Federated Machine Learning without Peeking into Your Data
Fu, F., Xue, H., Cheng, Y., Tao, Y., & Cui, B. (2022) · ACM SIGMOD International Conference on Management of Data (SIGMOD) · doi:10.1145/3514221.3526127 · Cited by 87
“, Party A owns the embedding table QA in the bottom model and obtains ∇EA in the backward propagation, which is done in several existing works [28, 36, 81].”
BackgroundMethodInfluentialFedSDG-FS: Efficient and Secure Feature Selection for Vertical Federated Learning
Li, A., Peng, H., Zhang, L., Huang, J., Guo, Q., Yu, H., & Liu, Y. (2023) · IEEE Conference on Computer Communications (INFOCOM) · doi:10.1109/infocom53939.2023.10228895 · Cited by 61
Practical Vertical Federated Learning With Unsupervised Representation Learning
Wu, Z., Li, Q., & He, B. (2022) · IEEE Transactions on Big Data · doi:10.1109/tbdata.2022.3180117 · Cited by 74
“Except for a study [32] that investigates the scenario where labels are shared across all the parties, most existing studies in vertical federated learning focus on another more practical scenario where only one party holds the labels.”
Adaptive Vertical Federated Learning on Unbalanced Features
Zhang, J., Guo, S., Qu, Z., Zeng, D., Wang, H., Liu, Q., & Zomaya, A. Y. (2022) · IEEE Transactions on Parallel and Distributed Systems · doi:10.1109/tpds.2022.3178443 · Cited by 60
“Feature-partitioned learning architectures have been developed for models including trees [30], linear and logistic regression [31], [32], [33], [34], [35], and neural networks [36], [37].”
BackgroundMethodEfficient Participant Contribution Evaluation for Horizontal and Vertical Federated Learning
Wang, J., Zhang, L., Li, A., You, X., & Cheng, H. (2022) · IEEE International Conference on Data Engineering (ICDE) · doi:10.1109/icde53745.2022.00073 · Cited by 61
Falcon: A Privacy-Preserving and Interpretable Vertical Federated Learning System
Wu, Y., Xing, N., Chen, G., Dinh, T. T. A., Luo, Z., Ooi, B. C., Xiao, X., & Zhang, M. (2023) · Proceedings of the VLDB Endowment · doi:10.14778/3603581.3603588 · Cited by 61
Towards communication-efficient vertical federated learning training via cache-enabled local updates
Fu, F., Miao, X., Jiang, J., Xue, H., & Cui, B. (2022) · Proceedings of the VLDB Endowment · doi:10.14778/3547305.3547316 · Cited by 51
Flexible Vertical Federated Learning With Heterogeneous Parties
Castiglia, T., Wang, S., & Patterson, S. (2023) · IEEE Transactions on Neural Networks and Learning Systems · doi:10.1109/tnnls.2023.3309701 · Cited by 56
“Many VFL algorithms assume parties have identical local training: each party uses the same local optimizer to update its model, typically standard stochastic gradient descent (SGD) [7], [8], [14], [15], [16], [17], [18].”
InfluentialVFL+: Low-Coupling Vertical Federated Learning With Privileged Information Paradigm
Dai, W., Cui, T., Zhang, T., & Chen, B. (2025) · IEEE Transactions on Emerging Topics in Computational Intelligence · doi:10.1109/tetci.2025.3543769 · Cited by 44
“Common scenarios include recommendation systems and user credit assessments, among others [16], [17].”
Gao, J., Wang, W., Liu, Z., Billah, M. F. R. M., & Campbell, B. (2021) · ACM Conference on Embedded Networked Sensor Systems (SenSys) · doi:10.1145/3485730.3493450 · Cited by 53
“In order to generate a robust load forecasting model, users intent to collaboratively train their collected energy data [11].”
BackgroundFederated matrix factorization with privacy guarantee
Li, Z., Ding, B., Zhang, C., Li, N., & Zhou, J. (2021) · Proceedings of the VLDB Endowment · doi:10.14778/3503585.3503598 · Cited by 55
“Existing work about other problems in the VFL setting includes learning tree models with secure multiparty computation techniques [32, 49] and training composed models [23].”
BackgroundBackdoor attacks and defenses in feature-partitioned collaborative learning
Liu, Y., Yi, Z., & Chen, T. (2020) · arXiv · doi:10.48550/arxiv.2007.03608 · Cited by 38
“…collaborative learning frameworks have been developed for models including trees (Cheng et al., 2019), linear and logistic regression (Yang et al., 2019b; Hardy et al., 2017; Gratton et al., 2018; Kikuchi et al., 2017; Hu et al., 2019a), and neural networks (Liu et al., 2018; Hu et al., 2019b).”
BackgroundHierarchical Federated Learning for Hybrid Data Partitioning Across Multitype Sensors
Su, L., & Lau, V. K. N. (2021) · IEEE Internet of Things Journal · doi:10.1109/jiot.2021.3051382 · Cited by 42
“Although [13] and [14] in vertical FL further utilize intermediate results exchange and the asynchronous update mechanism to address this issue, since now that there are multiple sensors in each type instead of one sensor that owns all indices of partial data as in the setting of vertical FL,…”
BackgroundMethodUllah, F., Moon, J., Naeem, H., & Jabbar, S. (2022) · The Journal of Supercomputing · doi:10.1007/s11227-022-04631-z · Cited by 41
Towards Efficient Synchronous Federated Training: A Survey on System Optimization Strategies
Jiang, Z., Wang, W., Li, B., & Yang, Q. (2022) · IEEE Transactions on Big Data · doi:10.1109/tbdata.2022.3177222 · Cited by 46
“cal FL and the community has not yet achieved a consensus on the use of any one [171], [172], [173], [174], much discussion in this survey is biased towards horizontal FL that has a widely acknowledged architecture (Section 2.”
BackgroundPrivColl: Practical Privacy-Preserving Collaborative Machine Learning
Zhang, Y., Bai, G., Li, X., Curtis, C., Chen, C., & Ko, R. K. L. (2020) · Lecture notes in computer science · doi:10.1007/978-3-030-58951-6_20 · Cited by 46
Collaboration in Participant-Centric Federated Learning: A Game-Theoretical Perspective
Huang, G., Chen, X., Ouyang, T., Ma, Q., Chen, L., & Zhang, J. (2022) · IEEE Transactions on Mobile Computing · doi:10.1109/tmc.2022.3194198 · Cited by 40
“Along a different line, in many application scenarios, FL training tasks are not generated by a central broker, but initiated by a group of users who themselves are also the participants in the FL [14], [15], in order to collectively train a common FL model.”
BackgroundA review of federated learning: taxonomy, privacy and future directions
Ratnayake, H., Chen, L., & Ding, X. (2023) · Journal of Intelligent Information Systems · doi:10.1007/s10844-023-00797-x · Cited by 38
“Privacy is one of the essential properties of federated learning, and a review of its privacy enhancements reveals that general privacy protection mechanisms can be adopted to protect the privacy of federated learning (Hu et al., 2019; Vaidya et al., 2013).”
Huang, C., Wang, L., & Han, X. G. (2023) · International World Wide Web Conference (WWW) · doi:10.1145/3543507.3583874 · Cited by 51
“Researchers have proposed diverse mechanisms, such as tree-based models [5, 7, 48] and neural networks [12, 19, 23] for VFL collaborations.”
BackgroundDecentralized Machine Learning Training: A Survey on Synchronization, Consolidation, and Topologies
Khan, Q. W., Khan, A. N., Rizwan, A., Ahmad, R., Khan, S., & Kim, D. (2023) · IEEE Access · doi:10.1109/access.2023.3284976 · Cited by 40
Learning Privately over Distributed Features: An ADMM Sharing Approach
Hu, Y., Liu, P., Kong, L., & Niu, D. (2019) · arXiv · doi:10.48550/arxiv.1907.07735 · Cited by 30
“A few recent studies [8, 9, 10, 11, 12, 13, 14] have revisited vertically partitioned features under the setting of distributed learning, which is motivated by the increase in feature dimensionality as well as the opportunity of cooperation between multiple parties that may hold different aspects of information about…”
BackgroundResultInfluentialSelf-citationDistributed artificial intelligence: Taxonomy, review, framework, and reference architecture
Janbi, N. F., Katib, I. A., & Mehmood, R. (2023) · Intelligent Systems with Applications · doi:10.1016/j.iswa.2023.200231 · Cited by 44
AsySQN: Faster Vertical Federated Learning Algorithms with Better Computation Resource Utilization
Zhang, Q., Gu, B., Deng, C., Gu, S., Bo, L., Pei, J., & Huang, H. (2021) · ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD) · doi:10.1145/3447548.3467169 · Cited by 44
“Such scenario is common in the industry applications of collaborative learning, such as medical study, financial risk, and targeted marketing Zhang et al. (2021); Gong et al. (2016); Yang et al. (2019b); Cheng et al. (2019); Liu et al. (2019b); Hu et al. (2019).”
BackgroundAgrEvader: Poisoning Membership Inference against Byzantine-robust Federated Learning
Zhang, Y., Bai, G., Chamikara, M. A. P., Ma, M., Shen, L., Wang, J., Nepal, ., Xue, M., Wang, L., & Liu, J. K. (2023) · International World Wide Web Conference (WWW) · doi:10.1145/3543507.3583542 · Cited by 49
Label synchronization for Hybrid Federated Learning in manufacturing and predictive maintenance
Rosero, R. L., Silva, C., Ribeiro, B., & Santos, B. F. (2024) · Journal of Intelligent Manufacturing · doi:10.1007/s10845-023-02298-8 · Cited by 38
“In this sense, various algorithms emerged: Federated Stochastic Block Coordinate Descent (FedBCD) (Novikova et al., 2022), Feature Distributed Machine Learning (FDML) (Hu et al., 2019), Heterogeneous Neural Networks (Het-eroNN) (Yang et al., 2019), SecureBoost (Cheng et al., 2019), Vertical Asynchronous FL (VAFL) (Che…”
Blockchain-Based Decentralized Federated Learning Method in Edge Computing Environment
Liu, S., Wang, X., Hui, L., & Wu, W. (2023) · Applied Sciences · doi:10.3390/app13031677 · Cited by 27
Differentially Private Vertical Federated Clustering
Li, Z., Wang, T., & Li, N. (2023) · Proceedings of the VLDB Endowment · doi:10.14778/3583140.3583146 · Cited by 34
Distributed Learning in the IoT–Edge–Cloud Continuum
Arzovs, A., Judvaitis, J., Nesenbergs, K., & Selavo, L. (2024) · Machine Learning and Knowledge Extraction · doi:10.3390/make6010015 · Cited by 36
“Furthermore, collaborative machine learning is used to describe DL in some articles, in which the FL and SL methods are defined as collaborative learning systems [48,49].”
MethodFeature Reconstruction Attacks and Countermeasures of DNN Training in Vertical Federated Learning
Ye, P., Jiang, Z., Wang, W., Li, B., & Li, B. (2024) · IEEE Transactions on Dependable and Secure Computing · doi:10.1109/tdsc.2024.3521451 · Cited by 34
“With its recent successes, deep neural networks (DNN) become particularly appealing in VFL [10], [11], [12].”
Efficient and Privacy-Preserving Feature Importance-Based Vertical Federated Learning
Li, A., Huang, J., Jia, J., Peng, H., Zhang, L., Tuan, L. A., Yu, H., & Li, X. (2023) · IEEE Transactions on Mobile Computing · doi:10.1109/tmc.2023.3333879 · Cited by 30
“Federated learning (FL) [1], [2], [3], [4], [5], [6] is an emerging collaborative machine learning paradigm which enables multiple data owners (a.k.a., FL clients) to jointly train a model by iteratively exchanging model parameters with an FL server without exposing local data.”
BackgroundAmigó, E., Gonzalo, J. A., & Mizzaro, S. (2021) · IEEE Transactions on Knowledge and Data Engineering · doi:10.1109/tkde.2021.3109823 · Cited by 23
“Some of them are related with methodology or general development frameworks [8], [13], [25], [29], [57].”
BackgroundImproving Privacy-Preserving Vertical Federated Learning by Efficient Communication with ADMM
Xie, C., Chen, P., Li, Q., Nourian, A., Zhang, C., & Li, B. (2024) · IEEE Conference on Secure and Trustworthy Machine Learning (SaTML) · doi:10.1109/satml59370.2024.00029 · Cited by 28
Accelerating Vertical Federated Learning
Cai, D., Fan, T. T., Kang, Y., Fan, L., Xu, M., Wang, S., & Yang, Q. (2022) · IEEE Transactions on Big Data · doi:10.1109/tbdata.2022.3192898 · Cited by 23
“al propose a novel vertical federated learning framework on the assumption of shared labels [13].”
BackgroundFEAST: A Communication-efficient Federated Feature Selection Framework for Relational Data
Fu, R., Wu, Y., Xu, Q., & Zhang, M. (2023) · Proceedings of the ACM on Management of Data · doi:10.1145/3588961 · Cited by 28
“Although VFL has gained traction in recent years, most works focus on model training and prediction [12, 27, 36, 37, 62] or privacy preservation [14, 32, 34, 60], instead of feature selection.”
BackgroundDecision Trees in Federated Learning: Current State and Future Opportunities
Heiyanthuduwage, S. R., Altas, I., Bewong, M., Islam, M. Z., & Deho, O. B. (2024) · IEEE Access · doi:10.1109/access.2024.3440998 · Cited by 21
“Pivot also addresses the limitation of the existing works where the system is designed considering the fact that an active party can share the label information in a plain text with the participating clients [114], [115].”
BackgroundZahir-Ismail, A.-T., & Shukla, R. M. (2025) · Scientific Reports · doi:10.1038/s41598-025-15993-8 · Cited by 21
Differentially Private Collaborative Coupling Learning for Recommender Systems
Zhang, Y., Bai, G., Zhong, M., Li, X., & Ko, R. K. L. (2020) · IEEE Intelligent Systems · doi:10.1109/mis.2020.3005930 · Cited by 22
“Pr1⁄2MðdÞ 2 S e Pr1⁄2Mðd0Þ 2 S þ d: (4)”
Achieving Model Fairness in Vertical Federated LearningPreprint
Liu, C., Fan, Z., Zhou, Z., Shi, Y., Pei, J., Chu, L., & Zhang, Y. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2109.08344 · Cited by 23
“The standard VFL methods are mostly designed to facilitate privacy preservation or/and efficient parallelized computation [7], [8], [26], [27], [28], [29], [30].”
MethodA mechanism design approach for multi-party machine learning
Chen, M., Liu, Y., Shen, W., Shen, Y., Tang, P., & Yang, Q. (2024) · Theoretical Computer Science · doi:10.1016/j.tcs.2024.114618 · Cited by 18
Continual Horizontal Federated Learning for Heterogeneous Data
Mori, J., Teranishi, I., & Furukawa, R. (2022) · International Joint Conference on Neural Networks (IJCNN) · doi:10.1109/ijcnn55064.2022.9892815 · Cited by 27
“There are VFL studies to handle various models such as linear/logistic regression [17]–[21], decision trees [22]–[26], neural networks [27]–[30], and other non-linear models [31], [32].”
Practical and General Backdoor Attacks Against Vertical Federated Learning
Xuan, Y., Chen, X., Zhao, Z., Tang, B., & Dong, Y. (2023) · Lecture notes in computer science · doi:10.1007/978-3-031-43415-0_24 · Cited by 23
Communication-Efficient Vertical Federated Learning via Compressed Error Feedback
Valdeira, P., Xavier, J., Soares, C., & Chi, Y. (2025) · IEEE Transactions on Signal Processing · doi:10.1109/tsp.2025.3540655 · Cited by 18
“This assumption, known as “relaxed protocol” [5], is sometimes made in VFL [14], [37], [38] and has applications, for example, in credit score prediction.”
Collaborative Distributed Machine Learning
Jin, D. S., Kannengießer, N., Rank, S., & Sunyaev, A. (2024) · ACM Computing Surveys · doi:10.1145/3704807 · Cited by 23
“Only a few real-world implementations of CDML systems exist.”
PFDRL: Personalized Federated Deep Reinforcement Learning for Residential Energy Management
Gao, J., Wang, W., Nikseresht, F., Rajan, V. G., & Campbell, B. (2023) · Proceedings of the International Conference on Parallel Processing · doi:10.1145/3605573.3605641 · Cited by 24
Zhang, Y., Sun, R., Shen, L., Bai, G., Xue, M., Meng, M. H., Li, X., Ko, R. K. L., & Nepal, . (2024) · International World Wide Web Conference (WWW) · doi:10.1145/3589334.3645545 · Cited by 27
Xia, W., Li, Y., Zhang, L., Wu, Z., & Yuan, X. (2023) · IEEE Transactions on Big Data · doi:10.1109/tbdata.2022.3231277 · Cited by 16
“Early research on VFL assumes that all the parties own the complete set of labels of data samples [4], [21], [22], [23] so that all the parties can calculate the loss function and update the model locally.”
BackgroundMa, B., Wu, J., Liu, W., Chiaraviglio, L., & Ming, X. (2020) · IEEE International Conference on Information Reuse and Integration (IRI) · doi:10.1109/iri49571.2020.00013 · Cited by 15
“The work [23] has shown that federated deep learning models or federated learning (FL) models have great potentials in learning effective representations and demonstrating cutting-edge performance in computer vision and natural language processing applications.”
BackgroundA Vertical Federated Learning Framework for Horizontally Partitioned LabelsPreprint
Xia, W., Li, Y., Zhang, L., Wu, Z., & Yuan, X. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2106.10056 · Cited by 23
Hsu, T.-C., & Lin, C. (2023) · Bioinformatics Advances · doi:10.1093/bioadv/vbac100 · Cited by 17
Multi-Participant Vertical Federated Learning Based Time Series Prediction
Yang, Y. T., Yang, G., Gao, Y., Zang, C., Chen, J., & Wang, Q. (2022) · International Conference on Computing and Artificial Intelligence (ICCAI) · doi:10.1145/3532213.3532238 · Cited by 18
“Besides, most of the VFL papers focus on the robustness to the attack[7, 11, 27, 28].”
BackgroundA systematic review on privacy-preserving distributed data mining
Sun, C., Ippel, L., Dekker, A. L. A. J., Dumontier, M., & van Soest, J. (2021) · Data Science · doi:10.3233/ds-210036 · Cited by 18
“In the past five years, use cases have been developed in healthcare [7, 88–90], finance [91], and technology companies [14, 92, 93] to examine different PPDDM methods.”
MethodA Differentially Private Blockchain-Based Approach for Vertical Federated Learning
Tran, L., Chari, S., Khan, M. S. I., Zachariah, A., Patterson, S., & Seneviratne, O. W. (2024) · IEEE International Conference on Decentralized Applications and Infrastructures (DAPPS) · doi:10.1109/dapps61106.2024.00020 · Cited by 16
“Vertical Federated Learning (VFL) [14], [15] is another form of FL where clients have the same set of samples, but each has a different feature set.”
BackgroundThe Internet of Things as a Deep Neural Network
Du, R., Magnússon, S., & Fischione, C. (2020) · IEEE Communications Magazine · doi:10.1109/mcom.001.2000015 · Cited by 13
Desirable Companion for Vertical Federated Learning: New Zeroth-Order Gradient Based Algorithm
Zhang, Q., Gu, B., Dang, Z., Deng, C. D., & Huang, H. (2021) · ACM International Conference on Information and Knowledge Management (CIKM) · doi:10.1145/3459637.3482249 · Cited by 20
“Most existing VFL frameworks adopt SGD methods Wan et al. (2007); Hu et al. (2019); Liu et al. (2019a); Gu et al. (2020).”
BackgroundMethodInfluentialEfficient Asynchronous Multi-Participant Vertical Federated Learning
Shi, H., Xu, Y., Jiang, Y., Yu, H., & Cui, L. (2022) · IEEE Transactions on Big Data · doi:10.1109/tbdata.2022.3201729 · Cited by 17
Privacy-Preserving Data Selection for Horizontal and Vertical Federated Learning
Zhang, L., Li, A., Peng, H., Han, F., Huang, F., & Li, X. (2024) · IEEE Transactions on Parallel and Distributed Systems · doi:10.1109/tpds.2024.3439709 · Cited by 13
“There are two main types of FL methods according to the distribution of data, horizontal federated learning (HFL) [1], [2] and vertical federated learning (VFL) [4], [5].”
BackgroundPrivacy-Preserving Vertical Federated Learning With Tensor Decomposition for Data Missing Features
Liao, T., Fu, L., Zhang, L., Yang, L., Chen, C., Ng, M. K., Huang, H., & Zheng, Z. (2025) · IEEE Transactions on Information Forensics and Security · doi:10.1109/tifs.2025.3552033 · Cited by 12
“[7] provides theoretical guarantees for the DP of the Gaussian mechanism, which are not as e ff ective as ( (cid:15), δ )-DP. Hu et al. [41] proposed FDML methods, which easily combine di ff erent privacy mechanisms to maintain a higher level of privacy.”
CampER: An Effective Framework for Privacy-Aware Deep Entity Resolution
Guo, Y., Chen, L., Zhou, Z., Zheng, B., Fang, Z., Zhang, Z., Mao, Y., & Gao, Y. (2023) · ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD) · doi:10.1145/3580305.3599266 · Cited by 12
Multi-party privacy-preserving decision tree training with a privileged party
Tong, Y., Feng, Q., Luo, M., & He, D. (2024) · Science China Information Sciences · doi:10.1007/s11432-023-4013-x · Cited by 12
Federated Recommendation Systems
Yang, Q. (2019) · IEEE International Conference on Big Data (BigData) · doi:10.1109/bigdata47090.2019.9005952 · Cited by 12
Zhang, J., Ge, S., Chang, T., & Luo, Z.-Q. T. (2022) · IEEE Transactions on Signal Processing · doi:10.1109/tsp.2022.3184772 · Cited by 13
“To evaluate the practical performance of the proposed algorithms, we consider the less studied distributed vertical learning problem [30], [31].”
BackgroundMethodTowards collaborative intelligence: routability estimation based on decentralized private data
Pan, J., Chang, C.-C., Xie, Z., Li, A., Tang, M., Zhang, T., Hu, J., & Chen, Y. (2022) · ACM/IEEE Design Automation Conference (DAC) · doi:10.1145/3489517.3530578 · Cited by 11
“1Engineering details including the implementation of FL framework [2, 11] and privacy concerns [19, 21] about the framework on both model and data have been well studied for general machine learning tasks.”
BackgroundFedlearn-Algo: A flexible open-source privacy-preserving machine learning platformPreprint
Liu, B., Tan, C., Wang, J., Zeng, T., Shan, H., Yao, H., Huang, H., Dai, P., Bo, L., & Chen, Y. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2107.04129 · Cited by 11
“According to the data partition differences, most of existing FL algorithms can be mainly categorized into horizontal FL algorithms and vertical FL algorithms (Yang et al., 2019).”
BackgroundTVFL: Tunable Vertical Federated Learning towards Communication-Efficient Model Serving
Wang, J., Zhang, L., Cheng, Y., Li, S., Zhang, H., Huang, D., & Xu, L. (2023) · IEEE Conference on Computer Communications (INFOCOM) · doi:10.1109/infocom53939.2023.10229061 · Cited by 12
VertiMRF: Differentially Private Vertical Federated Data Synthesis
Zhao, F., Li, Z., Ren, X., Ding, B., Yang, S., & Li, Y. (2024) · ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD) · doi:10.1145/3637528.3671771 · Cited by 10
Fairness-Driven Private Collaborative Machine Learning
Pessach, D., Tassa, T., & Shmueli, E. (2024) · ACM Transactions on Intelligent Systems and Technology · doi:10.1145/3639368 · Cited by 11
“These methods take two main approaches for preserving privacy.”
BackgroundMethodLai, J., Wang, T., Chen, C., Li, Y., & Zheng, Z. (2023) · ACM International Conference on Information and Knowledge Management (CIKM) · doi:10.1145/3583780.3615106 · Cited by 13
Task-Oriented Edge Networks: Decentralized Learning Over Wireless Fronthaul
Lee, H., & Kim, S. (2023) · IEEE Internet of Things Journal · doi:10.1109/jiot.2023.3347234 · Cited by 9
“As will be discussed in Section VI, (16) has been widely adopted in existing works [22], [26], [27], [29], [30].”
BackgroundMethodInfluentialComAI: Enabling Lightweight, Collaborative Intelligence by Retrofitting Vision DNNs
Jayarajah, K., Wanniarachchige, D., Abdelzaher, T., & Misra, A. (2022) · IEEE Conference on Computer Communications (INFOCOM) · doi:10.1109/infocom48880.2022.9796769 · Cited by 14
Adaptive federated learning scheme for recognition of malicious attacks in an IoT network
Chhikara, P., Tekchandani, R., & Kumar, N. (2023) · Computing · doi:10.1007/s00607-022-01146-6 · Cited by 9
FeaUn: Feature unlearning in vertical federated learning for IIoT against feature inference attacks
Lu, Z., Zhang, X., Li, T., Wang, Y., Li, G., & Liu, Z. (2025) · Neurocomputing · doi:10.1016/j.neucom.2025.131110 · Cited by 8
Loop-the-Loops: Fragmented Learning Over Networks for Constrained IoT Devices
Deb, P. K., Mukherjee, A., Singh, D., & Misra, S. (2022) · IEEE Transactions on Parallel and Distributed Systems · doi:10.1109/tpds.2022.3220221 · Cited by 9
“[10] proposed the division of Features instead of data points, to maintain feature privacy between nodes.”
BackgroundReFer: Retrieval-Enhanced Vertical Federated Recommendation for Full Set User Benefit
Li, W., Wang, Z., Wang, J., Xia, S., Zhu, J., Chen, M., Fan, J., Cheng, J., & Lei, J. (2024) · International ACM SIGIR Conference on Research and Development in Information Retrieval (SIGIR) · doi:10.1145/3626772.3657763 · Cited by 11
Toward Scalable and Privacy-preserving Deep Neural Network via Algorithmic-Cryptographic Co-design
Zhou, J. W., Zheng, L., Chen, C., Wang, Y., Zheng, X., Wu, B., Chen, C., Wang, L., & Yin, J. (2022) · ACM Transactions on Intelligent Systems and Technology · doi:10.1145/3501809 · Cited by 10
“Algorithmic methods build privacy-preserving DNN by spliting the computation graphs of DNN from an algorithmic perspective [21, 22, 28, 37, 44].”
MethodP$$^2$$CG: a privacy preserving collaborative graph neural network training framework
Miao, X., Zhang, W., Jiang, Y., Fu, F., Shao, Y., Chen, L., Tao, Y., Cao, G., & Cui, B. (2022) · The VLDB Journal · doi:10.1007/s00778-022-00768-8 · Cited by 9
Meng, F., Zhang, L., Chen, Y. A., & Wang, Y. (2024) · Proceedings on Engineering Sciences · doi:10.24874/pes06.02.017 · Cited by 10
Herb: Privacy-preserving Random Forest with Partially Homomorphic Encryption
Liao, Q., Cabral, B., Fernandes, J. P., & Lourenço, N. (2022) · International Joint Conference on Neural Networks (IJCNN) · doi:10.1109/ijcnn55064.2022.9892321 · Cited by 9
“Moreover, [12], [15] require a trusted third party for aggregating computation results which is often not practical to obtain [23] and can post additional privacy risks under the honest-but-curious assumption [24].”
BackgroundMethodResultInfluentialPrivacy Matters: Vertical Federated Linear Contextual Bandits for Privacy Protected Recommendation
Cao, Z., Liang, Z., Wu, B., Zhang, S., Li, H., Ouyang, W., Rong, Y., & Zhao, P. (2023) · ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD) · doi:10.1145/3580305.3599475 · Cited by 7
Preserving Privacy for Distributed Genome-Wide Analysis Against Identity Tracing Attacks
Zhang, Y., Bai, G., Li, X., Nepal, ., Grobler, M. M., Chen, C., & Ko, R. K. L. (2022) · IEEE Transactions on Dependable and Secure Computing · doi:10.1109/tdsc.2022.3186672 · Cited by 9
“DP is another methodology that constitutes a strong standard for privacy guarantees for algorithms on aggregate databases [5], [19], [34], [59], [61], [76], [80].”
BackgroundMethodResultInfluentialSecure Forward Aggregation for Vertical Federated Neural Networks
Cai, S., Chai, D., Liu, Y., Zhang, J., Jin, Y., Wang, L., Guo, K., & Chen, K. (2023) · Lecture notes in computer science · doi:10.1007/978-3-031-28996-5_9 · Cited by 12
Metadata
- Authors
- Hu, Y., Niu, D., Yang, J., & Zhou, S.
- Affiliations
- University of Alberta (CA)
- Tencent (China) (CN)
- Venue
- Proceedings of the 25th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining
- Publisher
- ACM
- Published
- 2019-07-25
- Pages
- 2232-2240
- OpenAlex
- W2951059495
- Cited by
- 163
- References
- 26
- Citations by year
- 2026: 7 · 2025: 21 · 2024: 29 · 2023: 30 · 2022: 33 · 2021: 28 · 2020: 12 · 2019: 3
- FWCI
- 7.27
- Type
- conference paper
- Language
- en
- Open access
- closed
- Locations
- ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD) (publishedVersion)
- Topics
- Physical Sciences › Computer Science › Artificial Intelligence › Privacy-Preserving Technologies in Data
- Physical Sciences › Computer Science › Computer Networks and Communications › Cryptography and Data Security
- Physical Sciences › Computer Science › Artificial Intelligence › Stochastic Gradient Optimization Techniques
- Keywords
- distributed learning, privacy-preserving machine learning, data ownership
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.