Federated Progressive Self-Distillation with Logits Calibration for Personalized IIoT Edge Intelligence

Wenqi Niu; Yingchao Wang

REVIEW 2 major objections 4 minor 28 references

Federated Progressive Self-Distillation with Logits Calibration for Personalized IIoT Edge Intelligence

T0 review · 2 major / 4 minor · reviewed 2026-08-12 · deepseek-v4-flash

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

Pith's one-line read A federated learning method counteracts knowledge forgetting by calibrating logits and distilling each client's own past outputs.

desk verdict A storage-light personalized FL recipe with big empirical gains, but the calibration component's theoretical motivation has a sign error that the authors need to fix. read the letter →

arxiv 2412.00410 v1 pith:FDHQKJB7 submitted 2024-11-30 cs.AI

Yingchao Wang , Wenqi Niu This is my paper · ORCID

classification cs.AI

keywords personalizedfederatedlearningknowledgeforgettingself-distillationlogitscalibrationNon-IIDdataIIoTedgeintelligencedistillation

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

Federated learning on non-IID edge data suffers from two kinds of forgetting: local training drifts away from the global model's general knowledge, and global aggregation erases each client's historical personalized knowledge. The paper proposes FedPSD, a client-side method that combines logits calibration with progressive self-distillation to prevent both kinds of forgetting. Each client's previous-round personalized model output becomes a soft label for the first local epoch, each previous local epoch serves as a virtual teacher for the next, and the local cross-entropy loss is reweighted by the class prior. On MNIST, CIFAR-10, and CIFAR-100 under sharding and LDA partitions, the paper reports higher Top-1 accuracy than eight federated learning baselines on both client and server models, plus faster convergence in communication rounds. The motivation is that personalized federated learning can then run on resource-constrained IIoT edge devices without storing an extra teacher model.

What carries the argument

The central mechanism is a pair of calibrated fusion labels and two client-side losses. The calibrated fusion label $H = \alpha P + (1-\alpha)Y$ with $\alpha = t/t_{\mathrm{total}}$ provides a smoothed teacher for the first local epoch, reviewing historical personalized knowledge. The progressive self-distillation loss $L_{KD} = \mathrm{KL}(H^{t,e-1}_k \| P^{t,e}_k)$ makes each epoch's output the teacher for the next, so global knowledge inherited from the server is passed along gradually. The calibrated cross-entropy loss uses $P_{\mathrm{calibrated}}(y|x) = P(y)e^{f_y(x)} / \sum_{y'} P(y')e^{f_{y'}(x)}$, and the total local loss is $L_{\mathrm{local}} = L_{CE} + L_{KD}$. The design keeps all modifications on the client side and requires no retained global teacher model.

What would settle it

Evaluate the decision rule implied by Eq (7) on two classes with logits $f_1=0$, $f_2=1$ and priors $P(1)=0.9$, $P(2)=0.1$: the calibrated probability in Eq (7) favors class 1, while the derived balanced rule $\arg\max_y(f_y(x) - \ln P(y))$ favors class 2, so the claimed equivalence is directly checkable. A decisive experiment would train on an imbalanced client with Eq (8) and with the explicit logit-subtraction loss, then compare per-class balanced accuracy; if the calibrated loss does not improve balanced accuracy over standard cross-entropy, the calibration claim is not doing the attributed work.

Watch

Extended reading notes

Core claim

FedPSD claims that the alternating pattern of local training and global aggregation in federated learning produces continual forgetting of both global generalization knowledge and historical personalized knowledge, and that both can be countered with changes confined to the client. For global knowledge, the local cross-entropy loss is replaced by a calibrated loss that reweights logits by the local class prior, and the model's output from the previous local epoch distills into the next epoch, so knowledge from the global model is inherited gradually. For personalized knowledge, the output of the client's personalized model from the previous communication round is fused with the ground-truth label into a calibrated fusion label that guides the initial epoch of local training. On MNIST, CIFAR-10, and CIFAR-100 under pathological sharding and LDA partitions, the paper reports consistent Top-1 accuracy gains over FedAvg, FedProx, FedCurv, FedNova, SCAFFOLD, MOON, FedNTD, and FedLMD, including CIFAR-10 S=2 client accuracy of 60.31 versus 42.42 for FedLMD and 42.66 for FedNTD, and faster convergence to FedAvg's final accuracy in fewer communication rounds.

Load-bearing premise

The load-bearing premise is that the calibrated loss of Eq (7), which multiplies the softmax by the class prior $P(y)$, realizes the balanced objective derived in Eq (6), $\arg\max_y(f_y(x) - \ln P(y))$; the two expressions have opposite signs, so if the premise gives way the method's theoretical motivation collapses.

Editorial extensions

If this is right

Edge clients can retain both global and personalized knowledge using only their own historical outputs, avoiding the storage cost of a separate teacher model.
Fewer communication rounds are needed to reach FedAvg's final accuracy; on CIFAR-10 sharding S=2, the client-side model reaches it in 10 rounds instead of 200.
The method's benefit grows as data heterogeneity increases, making it suited to IIoT deployments with skewed and missing classes.
Consistent gains across SimpleCNN, ResNet10, VGG11, and across 1–20 local epochs suggest the approach transfers across architectures and training budgets.

Reading between the lines

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

The derivation in Eqs (5)–(6) implies the balanced objective requires subtracting $\ln P(y)$ from logits, while Eq (7) adds it; if the intended calibration is the derived one, the method's theoretical justification would need a sign reversal, and the empirical gains might come from the progressive distillation rather than the calibrated loss.
A straightforward testable extension is to replace Eq (7) with the explicit logit adjustment $f_y(x) - \ln P(y)$ and compare balanced accuracy on the same partitions; this would isolate whether the calibration mechanism or the distillation schedule drives the reported gains.
The same calibrated-fusion idea could be applied to centralized class-imbalanced training, where past-epoch predictions could serve as smoothed teachers without any federated aggregation.
Storing one soft-label vector per client per round is cheap, but across many rounds memory could grow; averaging or compressing historical outputs is a natural follow-up that the paper does not explore.

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.

Desk Editor's Note

read the letter

Plain speaking: this is a useful empirical paper that overreaches in its theory. The proposed FedPSD combines three ingredients — progressive self-distillation where the previous local epoch's outputs teach the next, fusion labels from historical personalized models to seed the first epoch, and a prior-weighted local cross-entropy — and the combination is new. The storage overhead is genuinely low, which matters for the IIoT framing.

The experiments are the best part. The authors benchmark against eight baselines across MNIST, CIFAR-10, CIFAR-100, under both sharding and LDA, and report both client-side and server-side accuracy. The gains over FedLMD/FedNTD are large and consistent (e.g., CIFAR-10 S=2 client accuracy 60.31 vs 42.42). The ablations attribute gains to each component, though the names shift between text and table (GSD/RELogits vs RHPK/CLL) — a sloppiness that needs cleaning.

The soft spot is not soft: the theoretical motivation for the calibrated loss has a sign error. Eq. (6) correctly derives argmax(f_y(x) - ln P(y)) as the balanced objective. Eq. (7) then defines P_calibrated ∝ P(y) e^{f_y(x)}, which is argmax(f_y(x) + ln P(y)) — the opposite sign. The text says adding ln P(y) 'guides the model to prioritize locally frequent classes,' which directly contradicts the preceding derivation's goal of removing class-prior bias. The paper's central claim that CLL realizes the balanced objective is false as written. The ablation shows CLL accounts for a large chunk of the gain, so the method may work through some other reweighting effect, but the stated justification collapses.

The 'theoretical demonstration' in Section III is also not a proof; it is informal reasoning that the aggregation averages out personalized parameters, which is fine as intuition but not a theorem.

The empirical comparison is missing error bars, and several related methods cited in the related work (FedGKD, pFedSD, DKD-pFed) are not included as baselines. That is worth fixing.

Overall, the empirical core is plausible and the error is localized. I would send this to review with a request to fix the sign error or reframe CLL as a prior-weighted loss rather than a balanced-loss implementation, and to add error bars. The paper is likely to be a useful contribution to PFL if the authors make that repair.

Referee Report

2 major / 4 minor

Summary. The paper proposes FedPSD, a personalized federated learning method for IIoT edge clients, built from three client-side components: a logits-calibrated cross-entropy loss that rescales logits by local class priors, a progressive self-distillation loss in which the previous local epoch serves as the teacher for the next epoch, and calibrated fusion labels constructed from historical personalized model outputs that guide the first local epoch. The authors motivate the method with an analysis of global and personalized knowledge forgetting, and evaluate it on MNIST, CIFAR-10, and CIFAR-100 under pathological sharding and LDA partitions, reporting consistent Top-1 accuracy improvements on both client and server sides as well as reduced communication rounds to reach FedAvg-level accuracy.

Significance. The empirical study is extensive: eight baselines, three datasets, two non-IID partition families, ablations for each component, and robustness checks across network architectures, local epoch counts, and client participation rates. The experimental protocol is described in enough detail to be reproduced, and the reported gains are large and consistent. If the results hold, FedPSD is a practically useful contribution to personalized federated learning for resource-constrained edge devices. However, the central theoretical justification for the calibrated logits loss is internally inconsistent: the derivation in Eq. (6) and the implementation in Eq. (7) use opposite signs. This means the stated balanced-objective motivation for one of the three core components is invalid as written, and the empirical gains attributed to that component in Table VI currently lack the claimed theoretical explanation.

major comments (2)

[IV-C, Eqs. (5)-(8)] The derivation in Eq. (6) correctly obtains argmax_y(f_y(x) - ln P(y)) as the balanced objective that removes the class-prior bias. However, Eq. (7) defines P_calibrated(y|x) proportional to P(y)e^{f_y(x)}, whose argmax is argmax_y(f_y(x) + ln P(y)), because the normalization denominator is positive and independent of y. Thus the implemented loss in Eq. (8) has the opposite sign of the derived objective, and the text's statement that adding ln P(y) to the logits achieves the objective of Eq. (6) is false. This invalidates the theoretical motivation for the CLL component; the large CLL gains in Table VI cannot be explained by the balanced-objective argument as written. The authors should either change Eq. (7) to use P(y)^{-1}e^{f_y(x)} (equivalently, subtract ln P(y) from the logits) and rerun the affected experiments, or explicitly reframe CLL as a positive reweighting that prioritizes locally frequent classes and provide a separate justification for that design choice.
[III-A.1 and contribution list (Section I)] The claimed 'theoretical demonstration' that alternating aggregation and local training causes continual forgetting of global and personalized knowledge is not a proof. The decomposition w_t^k = w_global + w_local,k and the observation that the weighted average of personalized components may be small restate the effect of averaging but do not establish a mechanism by which a specific client's historical personalized knowledge is forgotten, nor do they quantify forgetting. The contribution bullet 'We theoretically demonstrate...' is therefore overstated. Please either provide a formal statement with assumptions and a quantitative bound or temper the claim to an analytical motivation.

minor comments (4)

[VI-A, Table VI] The ablation text refers to 'GSD' and 'RELogits' components, but the table columns are labeled RHPK, PSD, and CLL; additionally, the reported Δ values mix incremental gains over the previous configuration with gains compared to the baseline. Please align the terminology and state explicitly what each Δ measures.
[Figure 2 caption] The caption contains the phrase 'the sharing pathological sharding strategy'; this appears to be a typo for 'the sharding pathological sharding strategy' and should be corrected.
[V-D and Tables I-III] The results are reported as averages over three runs without standard deviations or per-seed ranges. Given the small number of runs, reporting variance information would substantially strengthen the reliability of the comparisons.
[IV-B-2, Eq. (3)] The within-round progressive self-distillation label H^{t,e-1}_k uses the same alpha from Eq. (2), which depends only on the communication round t, not on the local epoch e. Please clarify whether this is intentional, since the teacher here is the previous epoch's output within the same round rather than a historical personalized model.

Circularity Check

0 steps flagged · score 1.0 of 10

No significant circularity: the reported FedPSD accuracy gains come from standard benchmark comparisons with fixed hyperparameters, not from fitting the method to its own outputs or from a self-citation chain.

full rationale

The paper's central claim is an empirical comparison of FedPSD against baselines on MNIST, CIFAR-10, and CIFAR-100 under sharding and LDA partitions. The results are evaluated on held-out test sets, and the design parameters (e.g., the linear alpha schedule in Eq. (2), learning rate, batch size, number of local epochs) are fixed in advance and not tuned to the reported test accuracies. The logits-calibration component is motivated by a Bayes-derived balanced objective in Section IV-C; although the implementation in Eq. (7) appears to add rather than subtract ln P(y) relative to Eq. (6), that is a mathematical consistency error, not a circular dependency, because Eq. (7) is not fitted to the evaluation labels and the class prior P(y) is an input from the client's local data. The progressive self-distillation labels are constructed from the model's own previous outputs and ground-truth labels, which is the proposed training mechanism rather than a disguised form of the evaluation metric. The only self-citations (refs. [3] and [5]) are background survey citations and carry no load in the derivation of FedPSD or in the experimental conclusions. No derivation step reduces by construction to its own inputs, so circularity is minimal despite the noted local inconsistency in the CLL motivation.

Assumptions & free parameters 1 free parameters · 4 assumptions · 0 invented entities

The method relies on a few domain assumptions from prior literature (parameter decomposition, balanced objective) and a hand-picked alpha schedule. No new physical or mathematical entities are introduced. The sign error in the logits calibration is not an axiom but an internal error that weakens the derivation.

free parameters (1)

alpha schedule = t/t_total (linear growth from 0 to 1 over communication rounds)
Hand-chosen blending weight for historical personalized model outputs and ground-truth labels in the calibration fusion labels (Eq 1-2). The linear form is an ad hoc design choice; no search or justification is provided.

assumptions (4)

domain assumption Local model parameters can be decomposed as w_k = w_global + w_local,k.
Used in Section III-A to argue that averaging erodes personalized knowledge. Sourced from references [20], [21]; not validated for the models used in the experiments.
standard math Softmax outputs approximate P(y|x) as proportional to e^{f_y(x)}.
Standard softmax property invoked in Section IV-C to derive the balanced inference rule in Eq (5)-(6).
domain assumption The balanced posterior P_Bal(y|x) proportional to P(y|x)/P(y) is the appropriate personalization objective.
The paper asserts that removing the class prior yields fairer classification on imbalanced local data. This is a modeling choice, not proven, and is used to justify the calibrated loss.
domain assumption The previous epoch model's output, after linear calibration with ground-truth labels, is a reliable teacher for the next epoch.
Assumed throughout the progressive self-distillation loss in Section IV-B. No evidence is provided that such calibration prevents erroneous knowledge, despite the paper's claim that it does.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Federated Progressive Self-Distillation with Logits Calibration for Personalized IIoT Edge Intelligence." pith.science (2026). https://pith.science/paper/FDHQKJB7

@misc{pith2026241200410,
  author       = {Pith},
  title        = {Pith review of: Federated Progressive Self-Distillation with Logits Calibration for Personalized IIoT Edge Intelligence},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/FDHQKJB7}},
  note         = {Machine review of arXiv:2412.00410}
}

read the original abstract

Personalized Federated Learning (PFL) focuses on tailoring models to individual IIoT clients in federated learning by addressing data heterogeneity and diverse user needs. Although existing studies have proposed effective PFL solutions from various perspectives, they overlook the issue of forgetting both historical personalized knowledge and global generalized knowledge during local training on clients. Therefore, this study proposes a novel PFL method, Federated Progressive Self-Distillation (FedPSD), based on logits calibration and progressive self-distillation. We analyze the impact mechanism of client data distribution characteristics on personalized and global knowledge forgetting. To address the issue of global knowledge forgetting, we propose a logits calibration approach for the local training loss and design a progressive self-distillation strategy to facilitate the gradual inheritance of global knowledge, where the model outputs from the previous epoch serve as virtual teachers to guide the training of subsequent epochs. Moreover, to address personalized knowledge forgetting, we construct calibrated fusion labels by integrating historical personalized model outputs, which are then used as teacher model outputs to guide the initial epoch of local self-distillation, enabling rapid recall of personalized knowledge. Extensive experiments under various data heterogeneity scenarios demonstrate the effectiveness and superiority of the proposed FedPSD method.

Figures

Figures reproduced from arXiv: 2412.00410 by the authors.

**Figure 2.** Two distinct partitioning strategies were employed to segment the CIFAR-10 dataset, with visualizations of the distribution of different clients and [PITH_FULL_IMAGE:figures/full_fig_p006_2.png] view at source ↗

**Figure 3.** Top-1 accuracy during training for different methods, where [PITH_FULL_IMAGE:figures/full_fig_p008_3.png] view at source ↗

Figures from the paper (2 more)

**Figure 4.** Figure 4: Average client Top-1 accuracy at different epoch configurations. [PITH_FULL_IMAGE:figures/full_fig_p010_4.png]

**Figure 5.** Figure 5: Average client Top-1 accuracy at different clients configurations. [PITH_FULL_IMAGE:figures/full_fig_p010_5.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

28 extracted references · 11 canonical work pages

[1]

The internet of things in manufac- turing: Key issues and potential applications,

C. Yang, W. Shen, and X. Wang, “The internet of things in manufac- turing: Key issues and potential applications,” IEEE Systems, Man, and Cybernetics Magazine, vol. 4, no. 1, pp. 6–15, 2018

work page 2018
[2]

Industrial internet of things intelligence empowering smart manufacturing: A literature review,

Y . Hu, Q. Jia, Y . Yao, Y . Lee, M. Lee, C. Wang, X. Zhou, R. Xie, and F. R. Yu, “Industrial internet of things intelligence empowering smart manufacturing: A literature review,” IEEE Internet of Things Journal , 2024

work page 2024
[3]

Cloud-edge-device collaboration mechanisms of deep learning models for smart robots in mass personalization,

C. Yang, Y . Wang, S. Lan, L. Wang, W. Shen, and G. Q. Huang, “Cloud-edge-device collaboration mechanisms of deep learning models for smart robots in mass personalization,” Robotics and Computer- Integrated Manufacturing, vol. 77, p. 102351, 2022

2022
[4]

Federated learning for industrial internet of things in future industries,

D. C. Nguyen, M. Ding, P. N. Pathirana, A. Seneviratne, J. Li, D. Niyato, and H. V . Poor, “Federated learning for industrial internet of things in future industries,” IEEE Wireless Communications , vol. 28, no. 6, pp. 192–199, 2021

2021
[5]

End-edge-cloud collaborative computing for deep learning: A comprehensive survey,

Y . Wang, C. Yang, S. Lan, L. Zhu, and Y . Zhang, “End-edge-cloud collaborative computing for deep learning: A comprehensive survey,” IEEE Communications Surveys & Tutorials , 2024

2024
[6]

Federated optimization in heterogeneous networks,

T. Li, A. K. Sahu, M. Zaheer, M. Sanjabi, A. Talwalkar, and V . Smith, “Federated optimization in heterogeneous networks,” Proceedings of Machine learning and systems , vol. 2, pp. 429–450, 2020

2020
[7]

Towards personalized federated learning,

A. Z. Tan, H. Yu, L. Cui, and Q. Yang, “Towards personalized federated learning,” IEEE transactions on neural networks and learning systems , vol. 34, no. 12, pp. 9587–9603, 2022

2022
[8]

Rethinking personalized federated learning from knowledge perspective,

D. Yao, Z. Zhu, T. Liu, Z. Xu, and H. Jin, “Rethinking personalized federated learning from knowledge perspective,” in Proceedings of the 53rd International Conference on Parallel Processing , 2024, pp. 991– 1000

work page 2024

Show all 28 references

[9]

Preservation of the global knowledge by not-true distillation in federated learning,

G. Lee, M. Jeong, Y . Shin, S. Bae, and S.-Y . Yun, “Preservation of the global knowledge by not-true distillation in federated learning,” Ad- vances in Neural Information Processing Systems , vol. 35, pp. 38 461– 38 474, 2022

2022
[10]

Federated learning with label-masking distillation,

J. Lu, S. Li, K. Bao, P. Wang, Z. Qian, and S. Ge, “Federated learning with label-masking distillation,” inProceedings of the 31st ACM International Conference on Multimedia , 2023, pp. 222–232

2023
[11]

Learning critically: Selective self-distillation in federated learning on non-iid data,

Y . He, Y . Chen, X. Yang, H. Yu, Y .-H. Huang, and Y . Gu, “Learning critically: Selective self-distillation in federated learning on non-iid data,” IEEE Transactions on Big Data , 2022

2022
[12]

F ed gkd: Towards heterogeneous federated learning via global knowledge distillation,

D. Yao, W. Pan, Y . Dai, Y . Wan, X. Ding, C. Yu, H. Jin, Z. Xu, and L. Sun, “F ed gkd: Towards heterogeneous federated learning via global knowledge distillation,” IEEE Transactions on Computers , 2023

2023
[13]

Distilling the knowledge in a neural network,

G. Hinton, O. Vinyals, and J. Dean, “Distilling the knowledge in a neural network,” arXiv preprint arXiv:1503.02531 , 2015. 11

2015 arXiv
[14]

Learn from others and be yourself in heterogeneous federated learning,

W. Huang, M. Ye, and B. Du, “Learn from others and be yourself in heterogeneous federated learning,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2022, pp. 10 143–10 153

2022
[15]

Personal- ized edge intelligence via federated self-knowledge distillation,

H. Jin, D. Bai, D. Yao, Y . Dai, L. Gu, C. Yu, and L. Sun, “Personal- ized edge intelligence via federated self-knowledge distillation,” IEEE Transactions on Parallel and Distributed Systems , vol. 34, no. 2, pp. 567–580, 2023

2023
[16]

Dkd-pfed: A novel framework for personalized federated learning via decoupling knowledge distillation and feature decorrelation,

L. Su, D. Wang, and J. Zhu, “Dkd-pfed: A novel framework for personalized federated learning via decoupling knowledge distillation and feature decorrelation,” Expert Systems with Applications , vol. 259, p. 125336, 2025

2025
[17]

Overcoming forgetting in federated learning on non-iid data,

N. Shoham, T. Avidor, A. Keren, N. Israel, D. Benditkis, L. Mor-Yosef, and I. Zeitak, “Overcoming forgetting in federated learning on non-iid data,” arXiv preprint arXiv:1910.07796 , 2019

1910 arXiv
[18]

Acceleration of federated learning with alleviated forgetting in local training,

C. Xu, Z. Hong, M. Huang, and T. Jiang, “Acceleration of federated learning with alleviated forgetting in local training,” in The Tenth International Conference on Learning Representations, ICLR 2022, Virtual Event, April 25-29, 2022 , 2022

2022
[19]

Decoupled knowledge distillation,

B. Zhao, Q. Cui, R. Song, Y . Qiu, and J. Liang, “Decoupled knowledge distillation,” in Proceedings of the IEEE/CVF Conference on computer vision and pattern recognition , 2022, pp. 11 953–11 962

2022
[20]

Exploiting shared representations for personalized federated learning,

L. Collins, H. Hassani, A. Mokhtari, and S. Shakkottai, “Exploiting shared representations for personalized federated learning,” in Interna- tional conference on machine learning . PMLR, 2021, pp. 2089–2099

2021
[21]

Partialfed: Cross-domain personalized federated learning via partial initialization,

B. Sun, H. Huo, Y . Yang, and B. Bai, “Partialfed: Cross-domain personalized federated learning via partial initialization,” Advances in Neural Information Processing Systems , vol. 34, pp. 23 309–23 320, 2021

2021
[22]

A comprehensive survey of forgetting in deep learning beyond continual learning,

Z. Wang, E. Yang, L. Shen, and H. Huang, “A comprehensive survey of forgetting in deep learning beyond continual learning,” IEEE Trans- actions on Pattern Analysis and Machine Intelligence , 2024

2024
[23]

Gradient-based learning applied to document recognition,

Y . LeCun, L. Bottou, Y . Bengio, and P. Haffner, “Gradient-based learning applied to document recognition,” Proceedings of the IEEE , vol. 86, no. 11, pp. 2278–2324, 1998

1998
[24]

Learning multiple layers of features from tiny images,

A. Krizhevsky, G. Hinton et al. , “Learning multiple layers of features from tiny images,” 2009

2009
[25]

Communication-efficient learning of deep networks from decentralized data,

B. McMahan, E. Moore, D. Ramage, S. Hampson, and B. A. y Arcas, “Communication-efficient learning of deep networks from decentralized data,” in Artificial intelligence and statistics . PMLR, 2017, pp. 1273– 1282

2017
[26]

Model-contrastive federated learning,

Q. Li, B. He, and D. Song, “Model-contrastive federated learning,” in Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2021, pp. 10 713–10 722

2021
[27]

Tackling the objective inconsistency problem in heterogeneous federated optimiza- tion,

J. Wang, Q. Liu, H. Liang, G. Joshi, and H. V . Poor, “Tackling the objective inconsistency problem in heterogeneous federated optimiza- tion,” Advances in neural information processing systems , vol. 33, pp. 7611–7623, 2020

2020
[28]

Scaffold: Stochastic controlled averaging for federated learn- ing,

S. P. Karimireddy, S. Kale, M. Mohri, S. Reddi, S. Stich, and A. T. Suresh, “Scaffold: Stochastic controlled averaging for federated learn- ing,” in International conference on machine learning . PMLR, 2020, pp. 5132–5143

2020

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.

[1] [1]

The internet of things in manufac- turing: Key issues and potential applications,

C. Yang, W. Shen, and X. Wang, “The internet of things in manufac- turing: Key issues and potential applications,” IEEE Systems, Man, and Cybernetics Magazine, vol. 4, no. 1, pp. 6–15, 2018

work page 2018

[2] [2]

Industrial internet of things intelligence empowering smart manufacturing: A literature review,

Y . Hu, Q. Jia, Y . Yao, Y . Lee, M. Lee, C. Wang, X. Zhou, R. Xie, and F. R. Yu, “Industrial internet of things intelligence empowering smart manufacturing: A literature review,” IEEE Internet of Things Journal , 2024

work page 2024

[3] [3]

Cloud-edge-device collaboration mechanisms of deep learning models for smart robots in mass personalization,

C. Yang, Y . Wang, S. Lan, L. Wang, W. Shen, and G. Q. Huang, “Cloud-edge-device collaboration mechanisms of deep learning models for smart robots in mass personalization,” Robotics and Computer- Integrated Manufacturing, vol. 77, p. 102351, 2022

2022

[4] [4]

Federated learning for industrial internet of things in future industries,

D. C. Nguyen, M. Ding, P. N. Pathirana, A. Seneviratne, J. Li, D. Niyato, and H. V . Poor, “Federated learning for industrial internet of things in future industries,” IEEE Wireless Communications , vol. 28, no. 6, pp. 192–199, 2021

2021

[5] [5]

End-edge-cloud collaborative computing for deep learning: A comprehensive survey,

Y . Wang, C. Yang, S. Lan, L. Zhu, and Y . Zhang, “End-edge-cloud collaborative computing for deep learning: A comprehensive survey,” IEEE Communications Surveys & Tutorials , 2024

2024

[6] [6]

Federated optimization in heterogeneous networks,

T. Li, A. K. Sahu, M. Zaheer, M. Sanjabi, A. Talwalkar, and V . Smith, “Federated optimization in heterogeneous networks,” Proceedings of Machine learning and systems , vol. 2, pp. 429–450, 2020

2020

[7] [7]

Towards personalized federated learning,

A. Z. Tan, H. Yu, L. Cui, and Q. Yang, “Towards personalized federated learning,” IEEE transactions on neural networks and learning systems , vol. 34, no. 12, pp. 9587–9603, 2022

2022

[8] [8]

Rethinking personalized federated learning from knowledge perspective,

D. Yao, Z. Zhu, T. Liu, Z. Xu, and H. Jin, “Rethinking personalized federated learning from knowledge perspective,” in Proceedings of the 53rd International Conference on Parallel Processing , 2024, pp. 991– 1000

work page 2024