Pith. sign in

REVIEW 3 major objections 5 minor 71 references

UA-PDFL: A Personalized Approach for Decentralized Federated Learning

T0 review · 3 major / 5 minor · reviewed 2026-08-11 · deepseek-v4-flash

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

Pith's one-line read UA-PDFL claims adaptive personalization guided by a fixed unit tensor makes decentralized federated learning robust to data skew without public data or a central server.

desk verdict A practical DFL personalization method with a genuinely new unit-representation trick and broad experiments, but the load-bearing proxy is under-validated and the convergence proof does not analyze the actual algorithm. read the letter →

arxiv 2412.11674 v1 pith:UGSU3ORG submitted 2024-12-16 cs.LG cs.AI

Hangyu Zhu , Yuxiang Fan , Zhenping Xie This is my paper · ORCID
classification cs.LGcs.AI
keywords decentralizedfederatedlearningpersonalizedunitrepresentationnon-IIDdataclient-wisedropoutlayer-wisepersonalizationheterogeneity
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

This paper tries to make decentralized federated learning work when client data are unevenly distributed, without a central server and without any public data. It proposes UA-PDFL, in which every client feeds the same fixed tensor, all entries equal to one, through its local model and treats the output probability vector as a "unit representation" of that client's data distribution. The Jensen-Shannon divergence between these vectors decides which clients are similar enough to aggregate, and the number of shared feature-extractor layers is adapted per client rather than fixed. If the claim holds, a serverless federation can tune its personalization strength to the actual degree of data skew and stay accurate where fixed-personalization baselines falter.

What carries the argument

The load-bearing object is the unit representation $I_m = \sigma(f_m(w_m; X_{\mathrm{unit}}))$: the softmax output each client's model produces on the fixed unit tensor $X_{\mathrm{unit}}$, all entries equal to one. Pairwise Jensen-Shannon divergence $\mathrm{Div}(i,j)$ computed from these vectors drives every design choice: whether to drop clients when all divergences fall below threshold $th_I$, which classifiers to fuse, and which feature extractors to aggregate. A second auxiliary representation $I^{\mathrm{Aux}}_m = g_m(X_{\mathrm{unit}})$ from the feature extractor feeds a proximal regularizer $\mu\,\|I^{\mathrm{Aux}}_m - I^{\mathrm{Aux}}_{\mathrm{avg}}\|_2^2$ that pulls local extractors toward a shared feature space while classifiers stay personalized. The paper also justifies client-wise dropout as a Bayesian approximation, viewing each connected client as a neuron and the aggregation step as a linear combination.

What would settle it

Train two clients on the same data distribution with different random seeds or different architectures, then compute their unit representations and $\mathrm{Div}(i,j)$; if any such pair exceeds the paper's threshold $th_I$, the unit representation is not a faithful probe of true distribution similarity.

Watch

Extended reading notes

Core claim

The paper's central claim is that a single hand-set input, the unit tensor, can act as a distribution probe: feeding $X_{\mathrm{unit}}$ with all entries equal to $1$ to each client model and comparing the resulting softmax outputs with Jensen-Shannon divergence yields a metric $\mathrm{Div}(i,j)$ that ranks client-pair similarity in line with their true data distributions. On top of this metric the paper builds two mechanisms: client-wise dropout, which replaces a client's model with a randomly chosen connected client's model when all divergences are below a threshold, and layer-wise personalization, which aggregates feature extractors among similar clients while keeping each client's final model a weighted combination of similar peers' classifiers. A proximal term on auxiliary representations from the feature extractor steers local extractors toward a common feature space. The experiments on CIFAR-10, CIFAR-100, SVHN, and PathMNIST show UA-PDFL matching or beating five baselines and converging faster in most settings, with the largest gains on the hardest skewed task, CIFAR-100.

Load-bearing premise

The entire method rests on the assumption that a fixed all-ones input, pushed through each client's model, produces a unit representation whose pairwise Jensen-Shannon divergence reliably tracks how different the clients' real data distributions are.

Editorial extensions

If this is right

  • If UA-PDFL is right, decentralized federated learning no longer needs public data or a server to detect client similarity: a fixed dummy input and the Jensen-Shannon divergence between output vectors suffice.
  • Adaptive personalization depth should outperform fixed personalization layers on both ends of the skew spectrum, since client-wise dropout handles near-IID clients while layer-wise personalization handles strong non-IID.
  • The largest accuracy gains appear on the hardest heterogeneous tasks, CIFAR-100 and extreme PathMNIST non-IID, suggesting the mechanism is most valuable when client distributions overlap least.
  • Communication volume is not consistently reduced: the paper's own measurements show UA-PDFL communicates more than DFedAvgM and DisPFL, with savings from client-wise dropout appearing mainly under more uniform data.

Reading between the lines

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

  • Because $\mathrm{Div}$ lives in output-probability space, the method implicitly assumes all clients share the same label set and output layer; clients with different architectures or label spaces would need an agreed projection before the divergence is meaningful.
  • The unit representation is a single fixed probe, so its reliability may depend on model calibration; a natural test is whether averaging over many random fixed probes or using feature-map statistics makes the divergence more stable across architectures.
  • Client-wise dropout in the near-IID regime effectively turns training into mini-batch SGD on one random client per round, so for homogeneous data the method's benefit may reduce to communication savings rather than accuracy gains.
  • A straightforward extension would replace the hand-set threshold $th_I$ with an adaptive or per-client threshold derived from the divergence distribution, removing the method's most sensitive hyperparameter.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 5 minor

Summary. The paper proposes UA-PDFL, a decentralized federated learning method in which each client computes a 'unit representation' by feeding a fixed all-ones tensor through its local model, uses Jensen-Shannon divergence between these representations to decide client-wise dropout and layer-wise personalization, and adds an auxiliary feature-representation regularization term to the local objective. The method is evaluated on CIFAR-10, CIFAR-100, SVHN, and PathMNIST with three architectures and two Dirichlet heterogeneity levels, and is compared against five baselines plus an ablation. The paper claims to be the first to adaptively adjust the number of personalization layers via unit representations without relying on public data.

Significance. If the empirical results hold, UA-PDFL would offer a serverless FL approach that adapts personalization structure to the degree of data skew, avoids public data, and reduces communication overhead under relatively IID conditions. The experimental breadth is a genuine strength: four datasets, three architectures, two heterogeneity regimes, an extreme non-IID case, communication-cost measurements, and an ablation study. However, the load-bearing unit-representation proxy is not validated beyond a toy example and an incomplete ablation, and the convergence analysis in Section 4 does not model the actual algorithm. The empirical claims are plausible but not yet backed by a rigorous mechanism analysis, so the paper needs substantial revision before the central claims are fully supported.

major comments (3)
  1. [Section 3.2, Eqs. (5)-(7)] The central premise that Div(i,j) computed from unit representations on the all-ones tensor Xunit is a faithful proxy for the similarity of client data distributions is not established. Eq. (3) explicitly concedes that p(Dm) is not equal to p(σ(fm(wm; Xu))), and the subsequent statement that 'as Xunit ∈ Xu, it can inherently and accurately represent the data distribution differences across clients' is asserted without proof. Div(i,j) is a function of the model's output vector for a single out-of-distribution input; it can be insensitive to changes in conditional input distributions when label marginals coincide, and it can be driven by initialization, architecture, or out-of-distribution behavior. Since Algorithm 2 gates both dropout and layer-wise personalization on this metric, a failure of this premise makes the mechanism noise-driven. The toy experiment in Fig. 4 and the ablation in Fig. 12 only compare disjoint-label versus identical distributions. I would require a direct validation against a ground-truth distributional divergence, a sensitivity analysis over the entries aij, and at least one partial-overlap or same-marginal/different-domain scenario.
  2. [Section 4, Eq. (19) and Eq. (28)] The convergence analysis does not analyze the algorithm presented in Algorithms 1-2. Eq. (19) asserts Var(ξr) ≤ σ²/M 'by applying client dropout' with no derivation, and no connection is made to the threshold thI or to the dropout rule in Algorithm 2, which replaces the model with a random client only when all divergences are below thI and is inactive under heterogeneous data, precisely the setting where the empirical gains are claimed. The auxiliary representation term is dropped by asserting ∥I_Aux_m − I_Aux_avg∥² → 0 without proof, even though Eq. (12) makes it part of the local objective. Furthermore, the derivation tracks a single global parameter w, whereas the method maintains per-client feature extractors and classifiers that are aggregated differently. As written, Eq. (28) is a standard SGD convergence bound for a different algorithm; please either supply a proof for the actual UA-PDFL updates or clearly label Section 4 as a simplified informal analysis that does not make claims about the full method.
  3. [Section 5.5 and Algorithm 2] The ablation does not demonstrate that the unit representation, rather than the mechanism structure, is responsible for the reported gains. Full UA-PDFL is compared only against removing client-wise dropout or removing layer-wise personalization; there is no control that replaces Div(i,j) with random client selection or with a different similarity metric, and no sensitivity analysis for the threshold thI or the coefficient µ. Since thI determines when dropout replaces the local model and when classifiers are aggregated, the central claim that the method 'adaptively adjusts' personalization to data skew is not supported without showing how performance varies with thI and with the unit tensor entries aij. This is fixable by adding a small sensitivity study and a random-proxy baseline.
minor comments (5)
  1. [Algorithms 1 and 2] The computation of I_Aux_avg is inconsistent: Algorithm 1 line 17 appears to use only clients satisfying Div(m,i) < thI, while Algorithm 2 line 28 uses every client in Qi regardless of divergence; please harmonize the pseudocode and the surrounding text.
  2. [Section 5.1.4] The tables report mean±std but the number of independent runs is not stated; please provide the seed count and evaluation protocol so the reader can assess the stability of the comparisons.
  3. [Figure 11] The communication-cost figure reports values in units of 1e11 but the text does not explain how parameter transmission volume is counted (e.g., number of parameters × rounds × clients); please clarify the metric.
  4. [Throughout] There are numerous typos and English errors, such as 'an Unit representation' in Section 3, 'mantains' in Section 3.4, 'the the local loss' near Eq. (12), 'governs governs' in Section 5.1.2, and 'Leaning momentum' in Section 5.1.4; these should be corrected.
  5. [Section 5.2, Tables 1-2] The claim that UA-PDFL is 'superior or comparable' is based on final accuracy differences that are often within one standard deviation; the authors should avoid overstating small differences and ideally report pairwise significance tests or effect sizes.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the unit-representation divergence is an unproven validity assumption, not an input-to-output reduction, and the empirical results are evaluated on held-out test data.

full rationale

The paper's central empirical claim is validated against held-out test accuracy on CIFAR10/100, SVHN, and PathMNIST relative to external baselines (FedAvg, CriticalFL, FedPer, DisPFL, DFedAvgM), so no fitted parameter is renamed as a prediction. The unit-representation divergence Div(i,j) (Eqs. 5-7) is the main methodological risk: the paper admits in Eq. (3) that p(Dm) is not equal to p(σ(fm(wm;Xu))), and the assertion that Xunit 'inherently and accurately' represents distribution differences is supported only by a toy experiment and an ablation. This is an unproven validity assumption rather than circularity, since the divergence is not fitted to a target divergence and the toy/ablation compare it to known label distributions. The convergence analysis in Section 4.2 asserts that client dropout bounds the gradient-noise variance by σ^2/M and that the auxiliary term vanishes; these are unproven hypotheses, but the final bound is a standard consequence of those stated assumptions, not an equation equivalent to the algorithm's input by construction, and the theorem is not used to derive the experimental results. The only same-author citation, [62], is used to mirror a CNN architecture and is not load-bearing. No circular step meeting the evidence bar was found.

Assumptions & free parameters 5 free parameters · 5 assumptions · 2 invented entities

The paper's central mechanism rests on the unit-representation assumption that a model's output on a fixed unit tensor reflects the client's data distribution. This is an ad hoc domain assumption supported only by a toy experiment. The convergence proof further assumes the auxiliary regularization vanishes and that client-wise dropout reduces gradient noise to sigma^2/M, neither of which is derived from the algorithm. The main free parameters are the divergence threshold thI, the regularization coefficient mu, the queue size Ncom, the unit tensor entries, and the feature extractor split point.

free parameters (5)
  • divergence threshold thI = not reported
    Controls whether client-wise dropout or layer-wise personalization is triggered; no value or sensitivity analysis is given in the paper.
  • regularization coefficient mu = not reported
    Mixture weight for the auxiliary representation proximal term in Eq. (12); no value or tuning procedure is stated.
  • number of communicated clients Ncom = 5 (varied from 1 to 29 in Section 5.3)
    Sets queue size for peer-to-peer communication; the paper tests 1, 5, 15, 29 and uses 5 as default.
  • unit tensor entries aij = 1 (all entries)
    The fixed input used to compute unit representations; the paper does not study sensitivity to this choice.
  • feature extractor layer count = 2 (CNN), 60 (ResNet18), 28 (VGG11)
    Defines the split between general feature extractor and personalized classifier; chosen by hand without ablation.
assumptions (5)
  • standard math Local loss functions are L-smooth and mu-strongly convex (Assumption 1)
    Used in the Section 4 convergence analysis; not satisfied by neural networks in general.
  • standard math Bounded stochastic gradient variance (Assumption 2)
    Needed for the convergence bound; standard in FL theory.
  • domain assumption Bounded divergence between clients, Div(i,j) ≤ Delta (Assumption 3)
    Assumes all client divergences are bounded, even though the method is designed for settings where divergences vary widely.
  • ad hoc to paper Model inference on a fixed unit tensor reflects the client data distribution
    Core assumption in Section 3.2; only supported by a toy experiment, not a proof or sensitivity analysis.
  • ad hoc to paper Auxiliary representation term vanishes and client dropout bounds gradient noise variance by sigma^2/M
    Section 4.2 omits the auxiliary term by asserting it converges to zero and reduces variance to sigma^2/M without derivation from the algorithm.
invented entities (2)
  • Unit representation I_m
    purpose: A proxy for client data distribution computed as the output probabilities of a local model on a fixed unit tensor, used to compute pairwise divergence.
    Only validated in the paper's own toy experiment and ablation; no external benchmark or falsifiable prediction outside the framework.
  • Auxiliary representation I_aux_m
    purpose: Feature extractor output on the unit tensor, used in a proximal regularization term to align feature extractors across clients.
    Internal regularization device; no independent evidence of its effect beyond the paper's ablations.

how reviews work

0 comments
Cite this review

Pith. "Pith review of UA-PDFL: A Personalized Approach for Decentralized Federated Learning." pith.science (2026). https://pith.science/paper/UGSU3ORG

@misc{pith2026241211674,
  author       = {Pith},
  title        = {Pith review of: UA-PDFL: A Personalized Approach for Decentralized Federated Learning},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/UGSU3ORG}},
  note         = {Machine review of arXiv:2412.11674}
}
read the original abstract

Federated learning (FL) is a privacy preserving machine learning paradigm designed to collaboratively learn a global model without data leakage. Specifically, in a typical FL system, the central server solely functions as an coordinator to iteratively aggregate the collected local models trained by each client, potentially introducing single-point transmission bottleneck and security threats. To mitigate this issue, decentralized federated learning (DFL) has been proposed, where all participating clients engage in peer-to-peer communication without a central server. Nonetheless, DFL still suffers from training degradation as FL does due to the non-independent and identically distributed (non-IID) nature of client data. And incorporating personalization layers into DFL may be the most effective solutions to alleviate the side effects caused by non-IID data. Therefore, in this paper, we propose a novel unit representation aided personalized decentralized federated learning framework, named UA-PDFL, to deal with the non-IID challenge in DFL. By adaptively adjusting the level of personalization layers through the guidance of the unit representation, UA-PDFL is able to address the varying degrees of data skew. Based on this scheme, client-wise dropout and layer-wise personalization are proposed to further enhance the learning performance of DFL. Extensive experiments empirically prove the effectiveness of our proposed method.

Figures

Figures reproduced from arXiv: 2412.11674 by the authors.

Figure 1
Figure 1. A simple example of two different types of FL frameworks. (a) Centralized [PITH_FULL_IMAGE:figures/full_fig_p004_1.png] view at source ↗
Figure 2
Figure 2. Sequence diagram of DFL. Unlike traditional FL, the communication order could [PITH_FULL_IMAGE:figures/full_fig_p006_2.png] view at source ↗
Figure 3
Figure 3. An example of FL with personalization layers. [PITH_FULL_IMAGE:figures/full_fig_p008_3.png] view at source ↗
Figures from the paper (10 more)
Figure 4
Figure 4. Figure 4: The relationship between local data distribution and divergence metric. Firstly, [PITH_FULL_IMAGE:figures/full_fig_p012_4.png]
Figure 5
Figure 5. Figure 5: A simple example of client-wise dropout. (a) All the clients are connected for [PITH_FULL_IMAGE:figures/full_fig_p013_5.png]
Figure 6
Figure 6. Figure 6: An example of layer-wise personalization, where [PITH_FULL_IMAGE:figures/full_fig_p014_6.png]
Figure 7
Figure 7. Figure 7: Illustration of client data distributions with different [PITH_FULL_IMAGE:figures/full_fig_p022_7.png]
Figure 8
Figure 8. Figure 8: Illustration of client data with extreme cross-domain distribution. [PITH_FULL_IMAGE:figures/full_fig_p023_8.png]
Figure 9
Figure 9. Figure 9: The test accuracy over communication rounds with [PITH_FULL_IMAGE:figures/full_fig_p025_9.png]
Figure 10
Figure 10. Figure 10: The test accuracy over communication rounds with [PITH_FULL_IMAGE:figures/full_fig_p027_10.png]
Figure 11
Figure 11. Figure 11: Communication overhead across algorithms under varying data distributions [PITH_FULL_IMAGE:figures/full_fig_p030_11.png]
Figure 12
Figure 12. Figure 12: The divergence metric over training rounds. Each line represents the divergence [PITH_FULL_IMAGE:figures/full_fig_p031_12.png]
Figure 13
Figure 13. Figure 13: Test accuracy with or without layer-wise personalization or client-wise dropout, [PITH_FULL_IMAGE:figures/full_fig_p032_13.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

71 extracted references · 48 canonical work pages

  1. [1]

    Arivazhagan, M.G., Aggarwal, V., Singh, A.K., Choudhary, S.,

    work page

  2. [2]

    Incremental gradient, subgradient, and proximal methods for convex optimization: A survey

    Bertsekas, D.P., 2011. Incremental gradient, subgradient, and proximal methods for convex optimization: A survey

    work page 2011

  3. [3]

    Federated learning with hierar- chical clustering of local updates to improve training on non-iid data, in: 2020 International Joint Conference on Neural Networks (IJCNN), IEEE

    Briggs, C., Fan, Z., Andras, P., 2020. Federated learning with hierar- chical clustering of local updates to improve training on non-iid data, in: 2020 International Joint Conference on Neural Networks (IJCNN), IEEE. pp. 1–9. 32

    work page 2020

  4. [4]

    Communication- efficient and model-heterogeneous personalized federated learning via clustered knowledge transfer

    Cho, Y.J., Wang, J., Chirvolu, T., Joshi, G., 2023. Communication- efficient and model-heterogeneous personalized federated learning via clustered knowledge transfer. IEEE Journal of Selected Topics in Signal Processing 17, 234–247

    work page 2023

  5. [5]

    Dispfl: Towards communication-efficient personalized federated learning via decentral- ized sparse training, in: International conference on machine learning, PMLR

    Dai, R., Shen, L., He, F., Tian, X., Tao, D., 2022. Dispfl: Towards communication-efficient personalized federated learning via decentral- ized sparse training, in: International conference on machine learning, PMLR. pp. 4587–4604

    work page 2022

  6. [6]

    Personalized federated learning with theoretical guarantees: A model-agnostic meta-learning approach, in: Larochelle, H., Ranzato, M., Hadsell, R., Balcan, M., Lin, H

    Fallah, A., Mokhtari, A., Ozdaglar, A., 2020. Personalized federated learning with theoretical guarantees: A model-agnostic meta-learning approach, in: Larochelle, H., Ranzato, M., Hadsell, R., Balcan, M., Lin, H. (Eds.), Advances in Neural Information Processing Systems, Curran Associates, Inc.. pp. 3557–3568

    work page 2020

  7. [7]

    Federated learning for iot devices: En- hancing tinyml with on-board training

    Ficco, M., Guerriero, A., Milite, E., Palmieri, F., Pietrantuono, R., Russo, S., 2024. Federated learning for iot devices: En- hancing tinyml with on-board training. Information Fusion 104, 102189. URL: https://www.sciencedirect.com/science/article/ pii/S1566253523005055, doi: https://doi.org/10.1016/j.inffus. 2023.102189

    arXiv 2024

  8. [8]

    Dropout as a bayesian approximation: Insights and applications, in: deep learning workshop, ICML, p

    Gal, Y., Ghahramani, Z., 2015. Dropout as a bayesian approximation: Insights and applications, in: deep learning workshop, ICML, p. 2

    work page 2015

Show all 71 references
  1. [9]

    Dropout as a bayesian approxima- tion: Representing model uncertainty in deep learning, in: Balcan, M.F., Weinberger, K.Q

    Gal, Y., Ghahramani, Z., 2016. Dropout as a bayesian approxima- tion: Representing model uncertainty in deep learning, in: Balcan, M.F., Weinberger, K.Q. (Eds.), Proceedings of The 33rd International Con- ference on Machine Learning, PMLR, New York, New York, USA. pp. 1050–105...

    2016

  2. [10]

    Trusted decentralized federated learning, in: 2022 IEEE 19th Annual Consumer Communica- tions & Networking Conference (CCNC), IEEE

    Gholami, A., Torkzaban, N., Baras, J.S., 2022. Trusted decentralized federated learning, in: 2022 IEEE 19th Annual Consumer Communica- tions & Networking Conference (CCNC), IEEE. pp. 1–6

    2022

  3. [11]

    An efficient framework for clustered federated learning

    Ghosh, A., Chung, J., Yin, D., Ramchandran, K., 2022. An efficient framework for clustered federated learning. IEEE Transactions on In- formation Theory 68, 8076–8091. doi: 10.1109/TIT.2022.3192506. 33

    2022

  4. [12]

    Fedmcsa: Personalized federated learning via model components self-attention

    Guo, Q., Qi, Y., Qi, S., Wu, D., Li, Q., 2023. Fedmcsa: Personalized federated learning via model components self-attention. Neurocomput- ing 560, 126831. URL: https://www.sciencedirect.com/science/ article/pii/S0925231223009542, doi:https://doi.org/10.1016/j. neucom.2023.126831

    2023

  5. [13]

    Personalized federated learning: A unified framework and universal optimization techniques

    Hanzely, F., Zhao, B., et al., 2022. Personalized federated learning: A unified framework and universal optimization techniques. Transactions on Machine Learning Research

    2022

  6. [14]

    Deep residual learning for image recognition, in: Proceedings of the IEEE conference on computer vision and pattern recognition, pp

    He, K., Zhang, X., Ren, S., Sun, J., 2016. Deep residual learning for image recognition, in: Proceedings of the IEEE conference on computer vision and pattern recognition, pp. 770–778

    2016

  7. [15]

    Ran- domization is all you need: A privacy-preserving federated learn- ing framework for news recommendation

    Huang, X., Luo, Y., Liu, L., Zhao, W., Fu, S., 2023. Ran- domization is all you need: A privacy-preserving federated learn- ing framework for news recommendation. Information Sciences 637, 118943. URL: https://www.sciencedirect.com/science/ article/pii/S0020025523005121, doi:h...

    2023

  8. [16]

    A review of deep transfer learning and recent advancements

    Iman, M., Arabnia, H.R., Rasheed, K., 2023. A review of deep transfer learning and recent advancements. Technologies

    2023

  9. [17]

    A survey on federated learning for resource-constrained iot devices

    Imteaj, A., Thakker, U., Wang, S., Li, J., Amini, M.H., 2022. A survey on federated learning for resource-constrained iot devices. IEEE Internet of Things Journal 9, 1–24. doi: 10.1109/JIOT.2021.3095077

    2022

  10. [18]

    URL: https://www.mdpi.com/2227-7080/11/2/40, doi: 10.3390/ technologies11020040

  11. [19]

    Decentralized federated learning through proxy model sharing

    Kalra, S., Wen, J., Cresswell, J.C., Volkovs, M., Tizhoosh, H.R., 2023. Decentralized federated learning through proxy model sharing. Nature communications 14, 2899. 34

    2023

  12. [20]

    Personalized decentralized federated learning with knowledge distillation, in: ICC 2023 - IEEE Interna- tional Conference on Communications, pp

    Jeong, E., Kountouris, M., 2023. Personalized decentralized federated learning with knowledge distillation, in: ICC 2023 - IEEE Interna- tional Conference on Communications, pp. 1982–1987. doi: 10.1109/ ICC45041.2023.10279714

    2023

  13. [21]

    On information and sufficiency

    Kullback, S., Leibler, R.A., 1951. On information and sufficiency. The annals of mathematical statistics 22, 79–86

    1951

  14. [22]

    Learning multiple layers of features from tiny images

    Krizhevsky, A., Hinton, G., 2009. Learning multiple layers of features from tiny images. Master’s thesis, Department of Computer Science, University of Toronto

    2009

  15. [23]

    Decentralized federated learning via mutual knowledge transfer

    Li, C., Li, G., Varshney, P.K., 2021a. Decentralized federated learning via mutual knowledge transfer. IEEE Internet of Things Journal 9, 1136–1147

  16. [24]

    Fully de- centralized federated learning, in: Third workshop on bayesian deep learning (NeurIPS)

    Lalitha, A., Shekhar, S., Javidi, T., Koushanfar, F., 2018. Fully de- centralized federated learning, in: Third workshop on bayesian deep learning (NeurIPS)

    2018

  17. [25]

    Ditto: Fair and robust federated learning through personalization, in: International Conference on Machine Learning, PMLR

    Li, T., Hu, S., Beirami, A., Smith, V., 2021c. Ditto: Fair and robust federated learning through personalization, in: International Conference on Machine Learning, PMLR. pp. 6357–6368

  18. [26]

    Model-contrastive federated learning, in: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pp

    Li, Q., He, B., Song, D., 2021b. Model-contrastive federated learning, in: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pp. 10713–10722

  19. [27]

    Efficient decentral- ized optimization for edge-enabled smart manufacturing: A fed- erated learning-based framework

    Liu, H., Li, S., Li, W., Sun, W., 2024. Efficient decentral- ized optimization for edge-enabled smart manufacturing: A fed- erated learning-based framework. Future Generation Computer Systems URL: https://www.sciencedirect.com/science/article/ pii/S0167739X24001146, doi: https...

    2024 doi

  20. [28]

    Think locally, act globally: Federated learning with local and global representations

    Liang, P.P., Liu, T., Ziyin, L., Allen, N.B., Auerbach, R.P., Brent, D., Salakhutdinov, R., Morency, L.P., 2020. Think locally, act globally: Federated learning with local and global representations. arXiv preprint arXiv:2001.01523

    2020 arXiv

  21. [29]

    Multi- center federated learning: clients clustering for better personalization

    Long, G., Xie, M., Shen, T., Zhou, T., Wang, X., Jiang, J., 2023. Multi- center federated learning: clients clustering for better personalization. World Wide Web 26, 481–500

    2023

  22. [30]

    Fedcoin: A peer-to-peer payment system for federated learning, in: Federated learning: privacy and incentive

    Liu, Y., Ai, Z., Sun, S., Zhang, S., Liu, Z., Yu, H., 2020. Fedcoin: A peer-to-peer payment system for federated learning, in: Federated learning: privacy and incentive. Springer, pp. 125–138. 35

    2020

  23. [31]

    Decentralized federated learning: Fundamentals, state of the art, frameworks, trends, and challenges

    Mart ´ ınez Beltr´ an, E.T., P´ erez, M.Q., S´ anchez, P.M.S., Bernal, S.L., Bovet, G., P´ erez, M.G., P´ erez, G.M., Celdr´ an, A.H., 2023. Decentralized federated learning: Fundamentals, state of the art, frameworks, trends, and challenges. IEEE Communications Surveys & Tuto...

    2023

  24. [32]

    Like attracts like: Personalized federated learning in decentralized edge computing

    Ma, Z., Xu, Y., Xu, H., Liu, J., Xue, Y., 2022. Like attracts like: Personalized federated learning in decentralized edge computing. IEEE Transactions on Mobile Computing

    2022

  25. [33]

    The jensen- shannon divergence

    Men´ endez, M., Pardo, J., Pardo, L., Pardo, M., 1997. The jensen- shannon divergence. Journal of the Franklin Institute 334, 307–318

    1997

  26. [34]

    Flis: Clustered federated learning via inference similarity for non-iid data distribution

    Morafah, M., Vahidian, S., Wang, W., Lin, B., 2023a. Flis: Clustered federated learning via inference similarity for non-iid data distribution. IEEE Open Journal of the Computer Society 4, 109–120. doi: 10.1109/ OJCS.2023.3262203

    2023

  27. [35]

    McMahan, B., Moore, E., Ramage, D., Hampson, S., Arcas, B.A.y.,

  28. [36]

    A federated learning incen- tive mechanism in a non-monopoly market

    Na, S., Liang, Y., Yiu, S.M., 2024. A federated learning incen- tive mechanism in a non-monopoly market. Neurocomputing 586, 127630. URL: https://www.sciencedirect.com/science/article/ pii/S0925231224004016, doi: https://doi.org/10.1016/j.neucom. 2024.127630

    2024

  29. [37]

    Reading digits in natural images with unsupervised feature 36 learning, in: NIPS workshop on deep learning and unsupervised feature learning, Granada, Spain

    Netzer, Y., Wang, T., Coates, A., Bissacco, A., Wu, B., Ng, A.Y., et al., 2011. Reading digits in natural images with unsupervised feature 36 learning, in: NIPS workshop on deep learning and unsupervised feature learning, Granada, Spain. p. 7

    2011

  30. [38]

    Federated learning with partial model personalization, in: Chaudhuri, K., Jegelka, S., Song, L., Szepesvari, C., Niu, G., Sabato, S

    Pillutla, K., Malik, K., Mohamed, A.R., Rabbat, M., Sanjabi, M., Xiao, L., 2022. Federated learning with partial model personalization, in: Chaudhuri, K., Jegelka, S., Song, L., Szepesvari, C., Niu, G., Sabato, S. (Eds.), Proceedings of the 39th International Conference on Mac...

    2022

  31. [39]

    Flis: Clustered federated learning via inference similarity for non-iid data distribution

    Morafah, M., Vahidian, S., Wang, W., Lin, B., 2023b. Flis: Clustered federated learning via inference similarity for non-iid data distribution. IEEE Open Journal of the Computer Society 4, 109–120

  32. [40]

    Braintorrent: A peer-to-peer environment for decentralized federated learning

    Roy, A.G., Siddiqui, S., P¨ olsterl, S., Navab, N., Wachinger, C., 2019. Braintorrent: A peer-to-peer environment for decentralized federated learning. arXiv preprint arXiv:1905.06731

    2019 arXiv

  33. [41]

    Clustered federated learn- ing: Model-agnostic distributed multitask optimization under privacy constraints

    Sattler, F., M¨ uller, K.R., Samek, W., 2020. Clustered federated learn- ing: Model-agnostic distributed multitask optimization under privacy constraints. IEEE transactions on neural networks and learning sys- tems 32, 3710–3722

    2020

  34. [42]

    Ef- ficient and privacy-preserving online diagnosis scheme based on federated learning in e-healthcare system

    Shen, G., Fu, Z., Gui, Y., Susilo, W., Zhang, M., 2023. Ef- ficient and privacy-preserving online diagnosis scheme based on federated learning in e-healthcare system. Information Sciences 647, 119261. URL: https://www.sciencedirect.com/science/ article/pii/S0020025523008460, d...

    2023

  35. [43]

    Qu, Y., Dai, H., Zhuang, Y., Chen, J., Dong, C., Wu, F., Guo, S.,

  36. [44]

    Dropout: a simple way to prevent neural networks from overfitting

    Srivastava, N., Hinton, G., Krizhevsky, A., Sutskever, I., Salakhutdi- nov, R., 2014. Dropout: a simple way to prevent neural networks from overfitting. The journal of machine learning research 15, 1929–1958

    2014

  37. [45]

    Decentralized federated averaging

    Sun, T., Li, D., Wang, B., 2023. Decentralized federated averaging. IEEE Transactions on Pattern Analysis and Machine Intelligence 45, 4289–4301. doi: 10.1109/TPAMI.2022.3196503. 37

    2023

  38. [46]

    Gossipfl: A decentralized fed- erated learning framework with sparsified and adaptive communication

    Tang, Z., Shi, S., Li, B., Chu, X., 2022. Gossipfl: A decentralized fed- erated learning framework with sparsified and adaptive communication. IEEE Transactions on Parallel and Distributed Systems 34, 909–922

    2022

  39. [47]

    Towards personalized federated learning via heterogeneous model re- assembly

    Wang, J., Yang, X., Cui, S., Che, L., Lyu, L., Xu, D.D., Ma, F., 2024. Towards personalized federated learning via heterogeneous model re- assembly. Advances in Neural Information Processing Systems 36

    2024

  40. [48]

    Very deep convolutional networks for large-scale image recognition

    Simonyan, K., Zisserman, A., 2014. Very deep convolutional networks for large-scale image recognition. arXiv preprint arXiv:1409.1556

    2014 arXiv

  41. [49]

    Weng, J., Weng, J., Zhang, J., Li, M., Zhang, Y., Luo, W.,

  42. [50]

    Decentral and incentivized federated learning frameworks: A systematic literature review

    Witt, L., Heyer, M., Toyoda, K., Samek, W., Li, D., 2023. Decentral and incentivized federated learning frameworks: A systematic literature review. IEEE Internet of Things Journal 10, 3642–3663. doi: 10.1109/ JIOT.2022.3231363

    2023

  43. [51]

    Communication- efficient federated learning via knowledge distillation

    Wu, C., Wu, F., Lyu, L., Huang, Y., Xie, X., 2022. Communication- efficient federated learning via knowledge distillation. Nature communi- cations 13, 2032

    2022

  44. [52]

    Feddk: Improving cyclic knowledge distillation for personalized healthcare federated learning

    Xu, Y., Fan, H., 2023. Feddk: Improving cyclic knowledge distillation for personalized healthcare federated learning. IEEE Access 11, 72409– 72417. doi: 10.1109/ACCESS.2023.3294812

    2023

  45. [53]

    Personalized federated learning via heterogeneous mod- ular networks, in: 2022 IEEE International Conference on Data Mining (ICDM), pp

    Wang, T., Cheng, W., Luo, D., Yu, W., Ni, J., Tong, L., Chen, H., Zhang, X., 2022. Personalized federated learning via heterogeneous mod- ular networks, in: 2022 IEEE International Conference on Data Mining (ICDM), pp. 1197–1202. doi: 10.1109/ICDM54844.2022.00154

    2022

  46. [54]

    Yang, J., Shi, R., Wei, D., Liu, Z., Zhao, L., Ke, B., Pfister, H., Ni, B.,

  47. [55]

    IEEE Transactions on Dependable and Se- cure Computing 18, 2438–2455

    Deepchain: Auditable and privacy-preserving deep learning with blockchain-based incentive. IEEE Transactions on Dependable and Se- cure Computing 18, 2438–2455

  48. [56]

    Personalized and privacy-enhanced federated learning framework via knowledge distillation

    Yu, F., Wang, L., Zeng, B., Zhao, K., Yu, R., 2024. Personalized and privacy-enhanced federated learning framework via knowledge distillation. Neurocomputing 575, 127290. URL: https://www. sciencedirect.com/science/article/pii/S0925231224000614, doi:https://doi.org/10.1016/j.n...

    2024

  49. [57]

    Under- standing deep learning (still) requires rethinking generalization

    Zhang, C., Bengio, S., Hardt, M., Recht, B., Vinyals, O., 2021. Under- standing deep learning (still) requires rethinking generalization. Com- mun. ACM 64, 107–115. URL: https://doi.org/10.1145/3446776, doi:10.1145/3446776

    2021 doi

  50. [58]

    Fed- brain: A robust multi-site brain network analysis framework based on federated learning for brain disease diagnosis

    Zhang, C., Meng, X., Liu, Q., Wu, S., Wang, L., Ning, H., 2023. Fed- brain: A robust multi-site brain network analysis framework based on federated learning for brain disease diagnosis. Neurocomput- ing 559, 126791. URL: https://www.sciencedirect.com/science/ article/pii/S0925...

    2023

  51. [59]

    Yan, G., Wang, H., Yuan, X., Li, J., 2023. Criticalfl: A critical learn- ing periods augmented client selection framework for efficient federated learning, in: Proceedings of the 29th ACM SIGKDD Conference on Knowledge Discovery and Data Mining, pp. 2898–2907

    2023

  52. [60]

    Blockchain challenges and opportunities: A survey

    Zheng, Z., Xie, S., Dai, H.N., Chen, X., Wang, H., 2018. Blockchain challenges and opportunities: A survey. International journal of web and grid services 14, 352–375

    2018

  53. [61]

    Comavg: Robust decentralized federated learning with random committees

    Zhou, S., Huang, H., Li, R., Liu, J., Zheng, Z., 2023. Comavg: Robust decentralized federated learning with random committees. Computer Communications 211, 147–156

    2023

  54. [62]

    Robust clustered federated learning, in: International Conference on Database Systems for Advanced Applications, Springer

    Ye, T., Wei, S., Cui, J., Chen, C., Fu, Y., Gao, M., 2023. Robust clustered federated learning, in: International Conference on Database Systems for Advanced Applications, Springer. pp. 677–692

    2023

  55. [63]

    Federated learn- ing on non-iid data: A survey

    Zhu, H., Xu, J., Liu, S., Jin, Y., 2021. Federated learn- ing on non-iid data: A survey. Neurocomputing 465, 371–

    2021

  56. [66]

    Personalized feder- ated learning via variational bayesian inference, in: International Con- ference on Machine Learning, PMLR

    Zhang, X., Li, Y., Li, W., Guo, K., Shao, Y., 2022. Personalized feder- ated learning via variational bayesian inference, in: International Con- ference on Machine Learning, PMLR. pp. 26293–26310

    2022

  57. [69]

    Federated two-stage decoupling with adaptive personalization layers

    Zhu, H., Fan, Y., Xie, Z., 2024. Federated two-stage decoupling with adaptive personalization layers. Complex & Intelligent Sys- tems URL: https://doi.org/10.1007/s40747-024-01342-1 , doi:10. 1007/s40747-024-01342-1 . 39

    2024 doi

  58. [390]

    URL: https://www.sciencedirect.com/science/article/pii/ S0925231221013254, doi:https://doi.org/10.1016/j.neucom.2021. 07.098. 40

    2021 doi

  59. [2017]

    (Eds.), Proceedings of the 20th International Conference on Artificial Intelligence and Statistics, PMLR

    Communication-efficient learning of deep networks from de- centralized data, in: Singh, A., Zhu, J. (Eds.), Proceedings of the 20th International Conference on Artificial Intelligence and Statistics, PMLR. pp. 1273–1282. URL: https://proceedings.mlr.press/v54/ mcmahan17a.html

  60. [2019]

    arXiv preprint arXiv:1912.00818

    Federated learning with personalization layers. arXiv preprint arXiv:1912.00818

    1912 arXiv

  61. [2021]

    IEEE Network 35, 156–162

    Decentralized federated learning for uav networks: Architecture, challenges, and opportunities. IEEE Network 35, 156–162. doi:10.1109/ MNET.001.2100253

  62. [2023]

    Scientific Data 10, 41

    Medmnist v2-a large-scale lightweight benchmark for 2d and 3d biomedical image classification. Scientific Data 10, 41. 38

  63. [3013]

    doi: 10.1109/COMST.2023.3315746

    2023

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.