Federated Gaussian Mixture Models

Juan Carlos Andresen; Kuo-Yun Liang; Sophia Zhang Pettersson

REVIEW 3 major objections 4 minor 49 references

Federated Gaussian Mixture Models

T0 review · 3 major / 4 minor · reviewed 2026-08-07 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read FedGenGMM shows that a federated GMM can be aggregated in one communication round by sampling synthetic data from reweighted client components, matching non-federated training.

desk verdict Clean one-shot GMM aggregation idea, but the density-parity claim rests on a non-federated K choice and training-set likelihood; anomaly-detection results are more solid. read the letter →

arxiv 2506.01780 v1 pith:RYOH435F submitted 2025-06-02 cs.LG

Sophia Zhang Pettersson , Kuo-Yun Liang , Juan Carlos Andresen This is my paper

classification cs.LG

keywords federatedlearningGaussianmixturemodelsone-shotunsupervisedanomalydetectiondistributedEMstatisticalheterogeneitysyntheticdatageneration

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

FedGenGMM is a one-shot federated learning method for Gaussian mixture models (GMMs), a type of density model written as a weighted sum of Gaussians. Each client independently trains a local GMM with the standard expectation-maximization (EM) algorithm and sends only the model parameters and the local dataset size to the server; the server rescales the mixture weights by dataset size, samples a synthetic dataset from the pooled mixture, and trains a global GMM on those synthetic points. The paper claims that this single-round procedure learns the global data distribution about as well as a non-federated GMM trained on all the raw data, and better than the iterative distributed EM baselines tested. If correct, it gives edge deployments a communication-cheap way to do unsupervised density estimation and anomaly detection when clients see heterogeneous slices of the same underlying distribution.

What carries the argument

The carrying object is the synthetic dataset $S$ created on the server from the reweighted union of all client GMM components. Incoming component weights are rescaled by the factor $|D_c|/|D|$ before normalization, and the number of points drawn is $|S| = H\sum_c K_c$; the global model is then trained by EM on $S$ alone. This mechanism turns federated aggregation into a standard density-estimation problem and lets local models and the global model differ in component count and covariance structure.

What would settle it

Construct a federation in which two clients hold disjoint, well-separated modes of the global distribution (each client's local GMM has no component near the other's mode), run FedGenGMM, and compare the global model's average log-likelihood on held-out data from both modes against the non-federated EM benchmark. If the global model fits the union poorly while each local model fits its own mode well, the synthetic reweighted mixture has failed as a proxy for the global distribution, and the one-shot claim does not hold in that regime.

Watch

Extended reading notes

Core claim

The paper's central claim is that the generative property of Gaussian mixtures is enough to replace iterative federated parameter averaging in unsupervised learning. Given local GMMs $G_c$ trained by EM, the server forms a temporary pooled GMM by scaling each component weight by $|D_c|/|D|$, normalizing, drawing $|S|=H\sum_c K_c$ points from it, and running standard EM on $S$ to obtain the global model. Across six image, tabular, and time-series datasets, the resulting global model's average log-likelihood on the global distribution stays close to the non-federated EM benchmark and above the three distributed-EM variants under both Dirichlet and quantity-based heterogeneity; anomaly-detection area under the precision-recall curve is at least on par with the DEM baselines in most cases. The paper positions this as evidence that one-shot generative aggregation is competitive with, or better than, iterative distributed EM for federated GMM learning.

Load-bearing premise

The load-bearing premise is that a synthetic dataset sampled from the size-weighted combination of client Gaussian mixture components faithfully represents the true global data distribution; the server never sees raw data, so anything the local components fail to model is absent from the global model.

Editorial extensions

If this is right

With FedGenGMM, federated GMM training uses exactly one communication round per client, while the tested distributed-EM variants needed between roughly 3 and 41 rounds.
The global model's fit to the global distribution is stable across the tested heterogeneity levels, so feature-distribution skew alone does not degrade aggregation quality.
Client models can be deliberately smaller than the global model: on MNIST, Covertype, RWHAR, and VEHICLE, fewer local components still produce anomaly detection close to the 20-component non-federated benchmark.
The one-shot design gives a concrete communication-cost advantage for edge deployments, since clients send parameters once and then only receive the final global model.
The aggregation procedure is agnostic to how local models are trained, so alternative GMM training methods could replace EM client-side without changing the server protocol.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

Extending the paper: the same reweighted-sample aggregation could apply to any client-side generative density model that can emit samples, such as mixtures of Student-$t$ distributions or normalizing flows, provided the server knows local dataset sizes.
Extending the paper: the dependence on the hyperparameter $H$ (synthetic-data multiplier) is untested as a sensitivity axis; a natural extension is to measure global-model log-likelihood as $H$ shrinks toward the number of incoming components, where sampling noise should dominate.
Extending the paper: the assumption that pooled local components cover the global distribution could be checked before training the global model by comparing the likelihood of a held-out server-side sample under the pooled mixture and under the final global model.
Extending the paper: the single-round structure means a differential-privacy budget could be spent entirely on the one upload; the paper notes this possibility but does not implement or measure it.

Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

Dear colleague,

The short version: FedGenGMM is a simple one-shot scheme for aggregating locally trained GMMs by resampling synthetic data from the weighted mixture of client components, then training a global GMM on that synthetic set. It is clearly described, and the evaluation is broad. But the paper's headline claim—that it learns the global distribution on par with a non-federated model—is not actually established by the experiments. The two problems are exactly the ones flagged in the stress-test note: the global component count K is chosen by training non-federated GMMs on a validation split (Section 5.5), which leaks global information into the method, and the density-fitting metric is average log-likelihood on the training set (Section 5.6), which is an in-sample measure that structurally favors the non-federated benchmark trained on those exact points. Those two choices make the headline comparison a test of the method-with-oracle-K rather than the one-shot procedure as deployable.

What is genuinely new and good: the specific combination of unsupervised one-shot GMM aggregation with heterogeneous local component counts and an anomaly-detection evaluation. Prior work [2] has the synthetic-data-from-GMMs trick in a supervised feature-transfer setting, and [9] has a single-component variant; FedGenGMM extends it to full GMM density estimation and shows it works across image, tabular, and time-series data. The communication-cost table is straightforward and useful. The anomaly-detection experiments use holdout data and are therefore less affected by the in-sample issue; those results are the strongest part of the empirical section. The paper also explicitly admits that K selection would not be feasible in a federated setting, which is honest, but that honesty does not rescue the density-estimation comparison.

The soft spots beyond that: no code or data released, so the five-run averages are not independently checkable. The DEM baseline with init 2 gets a 100-point sample from the full training set, which gives it a mild global advantage while still being iterative; that is not fatal, but it blunts the comparison. The assumption that each client distribution is a mixture of shared global components (Eq. 1) is standard but not tested; if the local components do not line up, the reweighted mixture may be a poor proxy for p_D.

Who is this for? Someone building practical federated anomaly detection on edge fleets, or anyone looking for a baseline for one-shot unsupervised FL. It is not a field-reorganizing result, but it is a solid engineering contribution with a clear limitation. I would send it to a serious referee: the method is implementable, the communication gains are real, and the evaluation, though flawed, is extensive enough that a revision with a federated-compatible K-selection and a held-out log-likelihood comparison could make the claim stand. My own verdict is skeptical of the density-parity claim as currently supported, but the paper deserves a thorough review rather than a desk reject.

Best, [You]

Referee Report

3 major / 4 minor

Summary. The paper presents FedGenGMM, a one-shot federated method for training a global Gaussian mixture model. Each client fits a local GMM with EM and sends its parameters plus local dataset size; the server reweights component weights by dataset size (Eq. 4), constructs a synthetic dataset S by sampling from the reweighted mixture (Eq. 5), and trains a global GMM on S. The method is evaluated on six datasets (MNIST, Covertype, RWHAR, WADI, VEHICLE, SMD) under two heterogeneity schemes, against local models, three DEM baselines, and a non-federated EM benchmark. Metrics are average log-likelihood on the training set for density estimation and holdout AUC-PR for anomaly detection, plus communication rounds. The conclusions claim density-estimation parity with non-federated training and superiority over DEM with a single communication round.

Significance. The proposed aggregation is simple, practical, and flexible: it does not require identical local model structures, uses a single communication round, and can be paired with BIC-based local model selection. The anomaly-detection evaluation (Section 5.8) uses holdout data and is a genuine test of the method in a downstream task; the method performs at least on par with the DEM baselines in most settings, and the paper is candid about the infeasibility of its global-K selection in a federated setting. However, the headline density-estimation claim is not yet established: the evaluation in Sections 5.5-5.6 uses a non-federated validation set to choose K and reports training-set log-likelihood. Because these choices are load-bearing for the central claim, the paper needs a federated-compatible model-selection procedure and a held-out density-estimation comparison before the main conclusion can be accepted.

major comments (3)

[Section 5.5; Table 3; Figure 2] The global component count K (and, in most experiments, the local K_c) is chosen by training non-federated GMMs on a validation split and selecting a trade-off between model size and validation log-likelihood. The text explicitly states that this 'would not be feasible in a federated setting.' This means the density-estimation results in Figure 2 are conditional on global-distribution information that the one-shot federated procedure would not have. To support the conclusion that FedGenGMM learns the global distribution on par with a non-federated approach, the experiments should either use a federation-safe K selection (for example, local BIC per client with server-side aggregation of candidate models, evaluated with the chosen procedure) or the conclusions should be explicitly scoped to 'given a fixed, oracle-chosen K.'
[Section 5.6, Eq. (2), Figure 2] The density-estimation score is the average log-likelihood on the entire training dataset. This is an in-sample fit measure for every method, since the non-federated benchmark and DEM are trained on those exact points and the FedGenGMM pipeline's local models are fitted on the same data. It therefore does not measure how well any method approximates the population distribution p_D, which is what the 'learns the global data distribution' claim is about. The anomaly-detection experiments, which use a hold-out set, are not affected, but they do not test density-estimation parity. The authors should report average log-likelihood on a held-out test split for all methods, and ideally repeat the K-selection protocol using only training data.
[Section 4.1, Eq. (5), Section 5.5] The size of the synthetic dataset scales with H times the total number of local components, and H=100 is fixed in Section 5.5 without a sensitivity analysis. Since the global model is trained exclusively on S, the choice of H and the fidelity of the reweighted local mixture to p_D are first-order determinants of the method's behavior. At minimum, the paper should report how density-estimation and anomaly-detection results vary with H, or justify that the results are insensitive to it.

minor comments (4)

[Section 5.4 (DEM init 2)] The DEM init 2 baseline uses 100 data points drawn from the entire training dataset and sends them to the server; this gives that baseline access to raw global data and should be flagged when interpreting its communication cost and performance relative to FedGenGMM.
[Figures 2-5] The paper does not report statistical significance tests for the comparisons; with five runs and overlapping standard deviations, stronger claims such as 'better than the distributed EM methods' would benefit from paired tests or confidence intervals.
[Section 5.2] In the description of the artificial WADI classes, the notation C_m = 1(m-1)beta is unclear; the vector of ones should be written explicitly, for example as bold 1.
[Appendix A, Figure 6] The caption uses 'alpha' instead of the Greek symbol and repeats 'alpha' inconsistently; this should be cleaned up for consistency with the main text.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: FedGenGMM's global model is trained on a synthetic set sampled from reweighted local GMM parameters, and the reported density and anomaly comparisons are empirical results rather than consequences forced by the method's own definitions.

full rationale

The claimed derivation chain is: each client trains a local GMM by EM; the server reweights component weights by local dataset size (Eq. 4), forms a single GMM from all reweighted components, samples a synthetic dataset S of size proportional to the total number of components (Eq. 5), and then runs EM on S to obtain the global GMM. None of these steps defines the target quantity in terms of itself. The global model is not the reweighted union of local components; it is a freshly trained GMM fit to samples drawn from that union, so the reported global log-likelihood and anomaly-detection AUC-PR are empirical outcomes, not identities. The paper's only overlapping-author citation, [26], is used to attribute a data-partitioning scheme in Section 5.2; it is not load-bearing for the method or for any uniqueness claim. The main evaluation concerns are non-circular validity issues: Section 5.5 selects the global component count K by training non-federated GMMs on a validation split and lets K_c=K for FedGenGMM clients, and Section 5.6 reports average log-likelihood on the entire training dataset. These choices leak global information into the federated comparison and make the density-estimation headline harder to support, but they do not make the predicted likelihood a mathematical consequence of the fitted K or of the training data. The paper explicitly concedes that this way of determining K_c 'would not be feasible in a federated setting,' showing that the issue is a disclosed experimental limitation rather than a hidden equivalence. Since no quoted step reduces to its own inputs by construction, the appropriate circularity finding is zero.

Assumptions & free parameters 3 free parameters · 4 assumptions · 0 invented entities

The central claim rests on the mixture-of-distributions assumption for client data, on the fidelity of synthetic samples drawn from reweighted client components, and on the sufficiency of diagonal covariance matrices in the evaluated settings. Free parameters H and K are hand-set; no new entities are introduced.

free parameters (3)

H (synthetic data multiplier) = 100
Set in Section 5.5; controls synthetic dataset size |S| = H * sum K_c. Sensitivity to H is not studied, yet the global model quality depends on S being representative and sufficiently large.
K (global GMM components per dataset) = 10-30 depending on dataset (Table 3)
Selected by non-federated validation on the full training set, Section 5.5. This hand-set hyperparameter is unavailable in a true federated deployment.
K_c (local GMM components per client) = Equal to K in most experiments; 2-20 in constrained-client experiment
Local model complexity is a free design choice; the paper fixes it to K for comparability in most experiments and varies it in Section 5.8.

assumptions (4)

domain assumption Each client's local data distribution is a mixture of M shared underlying distributions p^(m), Eq. 1.
The aggregation pools components from different clients into one global mixture; if clients observe unrelated distributions, a single global GMM has no well-defined target.
domain assumption The synthetic dataset S, sampled from the reweighted union of client GMM components (Eqs. 4-5), is a faithful proxy for the global data distribution p_D.
The global model is trained exclusively on S; any mismatch between S and p_D directly degrades the density estimation and anomaly detection claims.
domain assumption Diagonal covariance matrices are sufficient to represent the global distributions of all six evaluated datasets.
All experiments use diagonal covariance (Section 5.5); the reported performance may not transfer to settings requiring full covariance.
standard math The EM algorithm, with k-means initialization and the stated convergence tolerance, produces adequate GMMs on local and synthetic data.
The method relies on standard EM behavior; no convergence guarantees are proved for the federated setting.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Federated Gaussian Mixture Models." pith.science (2026). https://pith.science/paper/RYOH435F

@misc{pith2026250601780,
  author       = {Pith},
  title        = {Pith review of: Federated Gaussian Mixture Models},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/RYOH435F}},
  note         = {Machine review of arXiv:2506.01780}
}

read the original abstract

This paper introduces FedGenGMM, a novel one-shot federated learning approach for Gaussian Mixture Models (GMM) tailored for unsupervised learning scenarios. In federated learning (FL), where multiple decentralized clients collaboratively train models without sharing raw data, significant challenges include statistical heterogeneity, high communication costs, and privacy concerns. FedGenGMM addresses these issues by allowing local GMM models, trained independently on client devices, to be aggregated through a single communication round. This approach leverages the generative property of GMMs, enabling the creation of a synthetic dataset on the server side to train a global model efficiently. Evaluation across diverse datasets covering image, tabular, and time series data demonstrates that FedGenGMM consistently achieves performance comparable to non-federated and iterative federated methods, even under significant data heterogeneity. Additionally, FedGenGMM significantly reduces communication overhead, maintains robust performance in anomaly detection tasks, and offers flexibility in local model complexities, making it particularly suitable for edge computing environments.

Figures

Figures reproduced from arXiv: 2506.01780 by the authors.

**Figure 2.** Resulting global model log likelihood mean for each dataset and method, using fixed [PITH_FULL_IMAGE:figures/full_fig_p011_2.png] view at source ↗

**Figure 3.** Anomaly detection AUC-PR mean for each dataset and method, using fixed [PITH_FULL_IMAGE:figures/full_fig_p014_3.png] view at source ↗

Figures from the paper (3 more)

**Figure 4.** Figure 4: Anomaly detection AUC-PR mean for each dataset and method, varying the number of clients from 20 to 320 while using [PITH_FULL_IMAGE:figures/full_fig_p015_4.png]

**Figure 5.** Figure 5: Anomaly detection AUC-PR mean for each dataset and method, when varying the number of GMM components ( [PITH_FULL_IMAGE:figures/full_fig_p016_5.png]

**Figure 6.** Figure 6: Resulting global model log likelihood mean for each dataset and method, using fixed [PITH_FULL_IMAGE:figures/full_fig_p019_6.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

49 extracted references · 30 canonical work pages

[2]

Mahdi Beitollahi, Alex Bie, Sobhan Hemati, Leo Maxime Brunswic, Xu Li, Xi Chen, and Guojun Zhang. 2024. Parametric Feature Transfer: One-shot Federated Learning with Foundation Models.arXiv e-prints, Article arXiv:2402.01862 (Feb. 2024), arXiv:2402.01862 pages. https: //doi.org/10.48550/arXiv.2402.01862 arXiv:2402.01862 [cs.LG]

arXiv doi:10.48550/arxiv.2402.01862 2024
[9]

Boyu Dong, Dong Chen, Yu Wu, Siliang Tang, and Yueting Zhuang. 2024. FADngs: Federated Learning for Anomaly Detection.IEEE Transactions on Neural Networks and Learning Systems(2024), 1–15. https://doi.org/10.1109/TNNLS.2024.3350660 Manuscript submitted to ACM Federated Gaussian Mixture Models 17

work page arXiv 2024
[1]

Chuadhry Mujeeb Ahmed, Venkata Reddy Palleti, and Aditya P. Mathur. 2017. WADI: a water distribution testbed for research in the design of secure cyber physical systems. InProceedings of the 3rd International Workshop on Cyber-Physical Systems for Smart Water Networks(Pittsburgh, Pennsylvania)(CySW ATER ’17). Association for Computing Machinery, New York,...

arXiv 2017
[3]

Jock Blackard. 1998. Covertype. UCI Machine Learning Repository. DOI: https://doi.org/10.24432/C50K5N. Manuscript submitted to ACM 16 Sophia Zhang Pettersson, Kuo-Yun Liang, and Juan Carlos Andresen (a) MNIST (b) Covertype (c) RWHAR (d) WADI (e) VEHICLE (f) SMD Fig. 5. Anomaly detection AUC-PR mean for each dataset and method, when varying the number of G...

doi:10.24432/c50k5n 1998
[4]

Criado, Fernando E

Marcos F. Criado, Fernando E. Casado, Roberto Iglesias, Carlos V. Regueiro, and Senén Barro. 2022. Non-IID data and Continual Learning processes in Federated Learning: A long road ahead.Information Fusion88 (2022), 263–280. https://doi.org/10.1016/j.inffus.2022.07.024

doi:10.1016/j.inffus.2022.07.024 2022
[5]

A. P. Dempster, N. M. Laird, and D. B. Rubin. 1977. Maximum likelihood from incomplete data via the EM algorithm.Journal of the Royal Statistical Society: Series B39 (1977), 1–38. http://web.mit.edu/6.435/www/Dempster77.pdf

1977
[6]

Li Deng. 2012. The MNIST Database of Handwritten Digit Images for Machine Learning Research [Best of the Web].IEEE Signal Processing Magazine 29, 6 (2012), 141–142. https://doi.org/10.1109/MSP.2012.2211477

arXiv 2012
[7]

Don Kurian Dennis, Tian Li, and Virginia Smith. 2021. Heterogeneity for the Win: One-Shot Federated Clustering. InInternational Conference on Machine Learning. https://api.semanticscholar.org/CorpusID:232075682

work page 2021

Show all 49 references

[8]

Xinrong Diao, Wei Yang, Shaowei Wang, Liusheng Huang, and Yang Xu. 2020. PrivGMM: Probability Density Estimation with Local Differential Privacy. InDatabase Systems for Advanced Applications. Springer International Publishing, Cham, 105–121

2020
[10]

Shaoming Duan, Chuanyi Liu, Peiyi Han, Xiaopeng Jin, Xinyi Zhang, Tianyu He, Hezhong Pan, and Xiayu Xiang. 2022. HT-Fed-GAN: Federated Generative Model for Decentralized Tabular Data Synthesis.Entropy25 (12 2022), 88. https://doi.org/10.3390/e25010088

2022 doi
[11]

Cynthia Dwork and Aaron Roth. 2014. The Algorithmic Foundations of Differential Privacy.Found. Trends Theor. Comput. Sci.9, 3–4 (Aug. 2014), 211–407. https://doi.org/10.1561/0400000042

2014 doi
[12]

Figueiredo and A.K

M.A.T. Figueiredo and A.K. Jain. 2002. Unsupervised learning of finite mixture models.IEEE Transactions on Pattern Analysis and Machine Intelligence 24, 3 (2002), 381–396. https://doi.org/10.1109/34.990138

2002 doi
[13]

Neel Guha, Ameet Talwalkar, and Virginia Smith. 2019. One-Shot Federated Learning.CoRRabs/1902.11175 (2019). arXiv:1902.11175 http: //arxiv.org/abs/1902.11175

2019 arXiv
[14]

Clare Elizabeth Heinbaugh, Emilio Luz-Ricca, and Huajie Shao. 2023. Data-Free One-Shot Federated Learning Under Very High Statistical Heterogeneity. InThe Eleventh International Conference on Learning Representations. https://openreview.net/forum?id=_hb4vM3jspB

2023
[15]

Chenxi Huang, Chaoyang Jiang, and Zhenghua Chen. 2023. Shuffled Differentially Private Federated Learning for Time Series Data Analytics. In 2023 IEEE 18th Conference on Industrial Electronics and Applications (ICIEA). 1023–1028. https://doi.org/10.1109/ICIEA58696.2023.10241529

2023
[16]

Salvatore Ingrassia and Roberto Rocci. 2011. Degeneracy of the EM algorithm for the MLE of multivariate Gaussian mixtures and dynamic constraints. Computational Statistics & Data Analysis55, 4 (2011), 1715–1725. https://doi.org/10.1016/j.csda.2010.10.026

2011 doi
[17]

Brendan McMahan, Brendan Avent, Aurélien Bellet, Mehdi Bennis, Arjun Nitin Bhagoji, Kallista Bonawit, Zachary Charles, Graham Cormode, Rachel Cummings, Rafael G

Peter Kairouz, H. Brendan McMahan, Brendan Avent, Aurélien Bellet, Mehdi Bennis, Arjun Nitin Bhagoji, Kallista Bonawit, Zachary Charles, Graham Cormode, Rachel Cummings, Rafael G. L. D’Oliveira, Hubert Eichner, Salim El Rouayheb, David Evans, Josh Gardner, Zachary Garrett, Adr...

2021
[18]

Siva Kasa and Vaibhav Rajan. 2023. Avoiding inferior clusterings with misspecified Gaussian mixture models.Scientific Reports13 (11 2023). https://doi.org/10.1038/s41598-023-44608-3

2023 doi
[19]

Anirudh Kasturi, Anish Reddy Ellore, and Chittaranjan Hota. 2020. Fusion Learning: A One Shot Federated Learning. InComputational Science – ICCS 2020, Valeria V. Krzhizhanovskaya, Gábor Závodszky, Michael H. Lees, Jack J. Dongarra, Peter M. A. Sloot, Sérgio Brissos, and João T...

2020
[20]

Leemaqz, Sharon X

Kaleb L. Leemaqz, Sharon X. Lee, and Geoffrey J. McLachlan. 2017. Corruption-Resistant Privacy Preserving Distributed EM Algorithm for Model-Based Clustering. In2017 IEEE Trustcom/BigDataSE/ICESS. 1082–1089. https://doi.org/10.1109/Trustcom/BigDataSE/ICESS.2017.356

2017 doi
[21]

Qinbin Li, Yiqun Diao, Quan Chen, and Bingsheng He. 2022. Federated Learning on Non-IID Data Silos: An Experimental Study. In2022 IEEE 38th International Conference on Data Engineering (ICDE). 965–978. https://doi.org/10.1109/ICDE53745.2022.00077

2022
[22]

Qiongxiu Li, Jaron Skovsted Gundersen, Katrine Tjell, Rafal Wisniewski, and Mads Græsbøll Christensen. 2022. Privacy-Preserving Distributed Expectation Maximization for Gaussian Mixture Model Using Subspace Perturbation. InICASSP 2022 - 2022 IEEE International Conference on Ac...

2022
[23]

Qinbin Li, Bingsheng He, and Dawn Xiaodong Song. 2020. Practical One-Shot Federated Learning for Cross-Silo Setting. InInternational Joint Conference on Artificial Intelligence. https://api.semanticscholar.org/CorpusID:235078788

2020
[24]

Tian Li, Anit Kumar Sahu, Manzil Zaheer, Maziar Sanjabi, Ameet Talwalkar, and Virginia Smith. 2020. Federated Optimization in Hetero- geneous Networks. InProceedings of the Third Conference on Machine Learning and Systems, MLSys 2020, Austin, TX, USA, March 2-4, 2020, Inderjit...

2020
[25]

Yan Li and Lei Li. 2009. A Novel Split and Merge EM Algorithm for Gaussian Mixture Model. In2009 Fifth International Conference on Natural Computation, Vol. 6. 479–483. https://doi.org/10.1109/ICNC.2009.625

2009 doi
[26]

Kuo-Yun Liang, Abhishek Srinivasan, and Juan Carlos Andresen. 2022. Modular Federated Learning. In2022 International Joint Conference on Neural Networks (IJCNN). IEEE, 1–8. https://doi.org/10.1109/IJCNN55064.2022.9892377

2022
[27]

Xiaodong Lin, Chris Clifton, and Michael Zhu. 2005. Privacy-preserving clustering with distributed EM mixture modeling.Knowl. Inf. Syst.8, 1 (July 2005), 68–81

2005
[28]

Erik William Lindskog-Münzing and Christian Prehofer. 2023. A Federated Learning Benchmark on Tabular Data: Comparing Tree-Based Models and Neural Networks. InInternational Conference on Fog and Mobile Edge Computing (FMEC). IEEE, 239–246. https://doi.org/10.1109/FMEC59375. 20...

2023
[29]

Abbass Madi, Oana Stan, Aurélien Mayoue, Arnaud Grivet-Sébert, Cédric Gouy-Pailler, and Renaud Sirdey. 2021. A Secure Federated Learning framework using Homomorphic Encryption and Verifiable Computing. In2021 Reconciling Data Analytics, Automation, Privacy, and Security: A Big...

2021
[30]

Othmane Marfoq, Giovanni Neglia, Aurélien Bellet, Laetitia Kameni, and Richard Vidal. 2021. Federated Multi-Task Learning under a Mixture of Distributions. InAdvances in Neural Information Processing Systems, M. Ranzato, A. Beygelzimer, Y. Dauphin, P.S. Liang, and J. Wortman V...

2021
[31]

Brendan McMahan, Eider Moore, Daniel Ramage, Seth Hampson, and Blaise Aguera y Arcas. 2017. Communication-Efficient Learning of Deep Networks from Decentralized Data. InProceedings of the 20th International Conference on Artificial Intelligence and Statistics (Proceedings of M...

2017
[32]

Kevin P. Murphy. 2012. Machine learning - a probabilistic perspective. InAdaptive computation and machine learning series. https://api.semanticscholar. org/CorpusID:17793133

2012
[33]

Nardi, L

M. Nardi, L. Valerio, and A. Passarella. 2022. Anomaly Detection Through Unsupervised Federated Learning. In2022 18th International Conference on Mobility, Sensing and Networking (MSN). IEEE Computer Society, Los Alamitos, CA, USA, 495–501. https://doi.org/10.1109/MSN57253.2022.00085

2022
[34]

Vibhor Pandhare, Jia Xiaodong, and Lee Jay. 2021. Collaborative Prognostics for Machine Fleets Using a Novel Federated Baseline Learner.Annual Conference of the PHM Society13 (11 2021). https://doi.org/10.36001/phmconf.2021.v13i1.2989

2021 doi
[35]

Felix Sattler, Klaus-Robert Müller, and Wojciech Samek. 2019. Clustered Federated Learning: Model-Agnostic Distributed Multitask Optimization Under Privacy Constraints.IEEE Transactions on Neural Networks and Learning Systems32 (2019), 3710–3722. https://api.semanticscholar.or...

2019
[36]

Felix Sattler, Simon Wiedemann, Klaus-Robert Müller, and Wojciech Samek. 2020. Robust and Communication-Efficient Federated Learning From Non-i.i.d. Data.IEEE Transactions on Neural Networks and Learning Systems31, 9 (2020), 3400–3413. https://doi.org/10.1109/TNNLS.2019.2944481

2020
[37]

Ya Su, Youjian Zhao, Chenhao Niu, Rong Liu, Wei Sun, and Dan Pei. 2019. Robust Anomaly Detection for Multivariate Time Series through Stochastic Recurrent Neural Network. InProceedings of the 25th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining(Anchora...

2019
[38]

Timo Sztyler and Heiner Stuckenschmidt. 2016. On-body localization of wearable devices: An investigation of position-aware activity recognition. In2016 IEEE International Conference on Pervasive Computing and Communications (PerCom). 1–9. https://doi.org/10.1109/PERCOM.2016.7456521

2016
[39]

Bram van Berlo, Aaqib Saeed, and Tanir Ozcelebi. 2020. Towards federated unsupervised representation learning. InProceedings of the Third ACM International Workshop on Edge Systems, Analytics and Networking(Heraklion, Greece)(EdgeSys ’20). Association for Computing Machinery, ...

2020
[40]

Vucovich, A

M. Vucovich, A. Tarcar, P. Rebelo, A. Rahman, D. Nandakumar, C. Redino, K. Choi, R. Schiller, S. Bhattacharya, B. Veeramani, A. West, and E. Bowen
[41]

Naibo Wang, Wenjie Feng, yuchen deng, Moming Duan, Fusheng Liu, and See-Kiong Ng. 2023. Data-Free Diversity-Based Ensemble Selection for One-Shot Federated Learning.Transactions on Machine Learning Research(2023). https://openreview.net/forum?id=ORMlg4g3mG

2023
[42]

Yang, Farhad Farokhi, Shi Jin, Tony Q

Kang Wei, Jun Li, Ming Ding, Chuan Ma, Howard H. Yang, Farhad Farokhi, Shi Jin, Tony Q. S. Quek, and H. Vincent Poor. 2020. Federated Learning With Differential Privacy: Algorithms and Performance Analysis.IEEE Transactions on Information Forensics and Security15 (2020), 3454–...

2020
[43]

Yuncheng Wu, Yao Wu, Hui Peng, Juru Zeng, Hong Chen, and Cuiping Li. 2016. Differentially private density estimation via Gaussian mixtures model. In2016 IEEE/ACM 24th International Symposium on Quality of Service (IWQoS). 1–6. https://doi.org/10.1109/IWQoS.2016.7590445

2016
[44]

Yue Wu, Shuaicheng Zhang, Wenchao Yu, Yanchi Liu, Quanquan Gu, Dawei Zhou, Haifeng Chen, and Wei Cheng. 2023. Personalized federated learning under mixture of distributions. InProceedings of the 40th International Conference on Machine Learning(Honolulu, Hawaii, USA)(ICML’23)....

2023
[45]

Fengda Zhang, Kun Kuang, Long Chen, Zhaoyang You, Tao Shen, Jun Xiao, Yin Zhang, Chao Wu, Yueting Zhuang, and Xiaolin Li. 2020. Federated unsupervised representation learning.Frontiers of Information Technology & Electronic Engineering24 (2020), 1181 – 1193. https://api.semant...

2020
[46]

Jie Zhang, Chen Chen, Bo Li, Lingjuan Lyu, Shuang Wu, Shouhong Ding, Chunhua Shen, and Chao Wu. 2024. DENSE: data-free one-shot federated learning. InProceedings of the 36th International Conference on Neural Information Processing Systems(New Orleans, LA, USA)(NIPS ’22). Curr...

2024
[47]

Yanlin Zhou, George Pu, Xiyao Ma, Xiaolin Li, and Dapeng Wu. 2020. Distilled One-Shot Federated Learning.CoRRabs/2009.07999 (2020). arXiv:2009.07999 https://arxiv.org/abs/2009.07999

2020 arXiv
[48]

Zhuangdi Zhu, Junyuan Hong, and Jiayu Zhou. 2021. Data-Free Knowledge Distillation for Heterogeneous Federated Learning.Proceedings of machine learning research139 (07 2021), 12878–12889. Manuscript submitted to ACM Federated Gaussian Mixture Models 19 A Full plots for estimat...

2021
[2023]

In2023 33rd International Telecommunication Networks and Applications Conference (ITNAC)

Anomaly Detection via Federated Learning. In2023 33rd International Telecommunication Networks and Applications Conference (ITNAC). IEEE Computer Society, Los Alamitos, CA, USA, 259–266. https://doi.org/10.1109/ITNAC59571.2023.10368517

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.

[1] [2]

Mahdi Beitollahi, Alex Bie, Sobhan Hemati, Leo Maxime Brunswic, Xu Li, Xi Chen, and Guojun Zhang. 2024. Parametric Feature Transfer: One-shot Federated Learning with Foundation Models.arXiv e-prints, Article arXiv:2402.01862 (Feb. 2024), arXiv:2402.01862 pages. https: //doi.org/10.48550/arXiv.2402.01862 arXiv:2402.01862 [cs.LG]

arXiv doi:10.48550/arxiv.2402.01862 2024

[2] [9]

Boyu Dong, Dong Chen, Yu Wu, Siliang Tang, and Yueting Zhuang. 2024. FADngs: Federated Learning for Anomaly Detection.IEEE Transactions on Neural Networks and Learning Systems(2024), 1–15. https://doi.org/10.1109/TNNLS.2024.3350660 Manuscript submitted to ACM Federated Gaussian Mixture Models 17

work page arXiv 2024

[3] [1]

Chuadhry Mujeeb Ahmed, Venkata Reddy Palleti, and Aditya P. Mathur. 2017. WADI: a water distribution testbed for research in the design of secure cyber physical systems. InProceedings of the 3rd International Workshop on Cyber-Physical Systems for Smart Water Networks(Pittsburgh, Pennsylvania)(CySW ATER ’17). Association for Computing Machinery, New York,...

arXiv 2017

[4] [3]

Jock Blackard. 1998. Covertype. UCI Machine Learning Repository. DOI: https://doi.org/10.24432/C50K5N. Manuscript submitted to ACM 16 Sophia Zhang Pettersson, Kuo-Yun Liang, and Juan Carlos Andresen (a) MNIST (b) Covertype (c) RWHAR (d) WADI (e) VEHICLE (f) SMD Fig. 5. Anomaly detection AUC-PR mean for each dataset and method, when varying the number of G...

doi:10.24432/c50k5n 1998

[5] [4]

Criado, Fernando E

Marcos F. Criado, Fernando E. Casado, Roberto Iglesias, Carlos V. Regueiro, and Senén Barro. 2022. Non-IID data and Continual Learning processes in Federated Learning: A long road ahead.Information Fusion88 (2022), 263–280. https://doi.org/10.1016/j.inffus.2022.07.024

doi:10.1016/j.inffus.2022.07.024 2022

[6] [5]

A. P. Dempster, N. M. Laird, and D. B. Rubin. 1977. Maximum likelihood from incomplete data via the EM algorithm.Journal of the Royal Statistical Society: Series B39 (1977), 1–38. http://web.mit.edu/6.435/www/Dempster77.pdf

1977

[7] [6]

Li Deng. 2012. The MNIST Database of Handwritten Digit Images for Machine Learning Research [Best of the Web].IEEE Signal Processing Magazine 29, 6 (2012), 141–142. https://doi.org/10.1109/MSP.2012.2211477

arXiv 2012

[8] [7]

Don Kurian Dennis, Tian Li, and Virginia Smith. 2021. Heterogeneity for the Win: One-Shot Federated Clustering. InInternational Conference on Machine Learning. https://api.semanticscholar.org/CorpusID:232075682

work page 2021