Pith. sign in

REVIEW 5 major objections 6 minor 42 references

Federated Unlearning Model Recovery in Data with Skewed Label Distributions

T0 review · 5 major / 6 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 After a skewed-class client is unlearned, the remaining clients can synthesize and denoise that class's data to restore model accuracy without touching the leaving client's data.

desk verdict A credible modest empirical result for federated unlearning recovery under label skew, but the method assumes every remaining client already has skewed-class seeds and the zero-sample case is untested. read the letter →

arxiv 2412.13466 v2 pith:K7IUSTJR submitted 2024-12-18 cs.LG

Xinrui Yu , Wenbin Pei , Bing Xue , Qiang Zhang This is my paper · ORCID
classification cs.LG
keywords federatedunlearninglabeldistributionskewmodelrecoverySMOTEautoencoderdensity-baseddenoisingimbalancedlearning
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

The paper tackles a blind spot in federated unlearning: when a client whose data is dominated by one class leaves and is unlearned, the global model's accuracy on that class collapses, and the remaining clients have too few of that class's samples to retrain the bias away. It proposes a recovery method, Imba-ULRc, in which each remaining client trains a local autoencoder, oversamples the skewed class in the latent space with SMOTE, removes low-density synthetic outliers, and then uses the enhanced local datasets for federated recovery training. The central claim is that this restores the unlearning model's performance on the skewed class and overall without ever accessing the leaving client's data. If true, it gives federated systems a way to honor unlearning requests without accepting a permanently degraded model whenever label distributions are skewed.

What carries the argument

The load-bearing machinery is a local autoencoder whose training objective combines a standard reconstruction loss with a reverse-order reconstruction loss: the decoder is fed encoded features in the reverse of the order they were produced, so the model learns to map latent variations back to the original data space. SMOTE then interpolates new latent points between a skewed-class sample and a neighboring encoded sample, and the decoder turns them into synthetic data. To keep that data clean, each generated point receives a density score from the distances to its k nearest same-class neighbors, and its density factor is the ratio of the neighbors' average density to its own; generated points whose factor exceeds the median are discarded. The surviving data are used for local gradient-descent recovery training, and the server aggregates the local models with data-size weights.

What would settle it

Run the recovery pipeline on a split where one remaining client has zero skewed-class samples: the SMOTE step for that client is undefined and the code would have no seed to interpolate from. A recovery result on such a split would require a fallback mechanism the paper does not specify; conversely, confirming that accuracy collapses only in that zero-sample case would localize the method's boundary.

Watch

Extended reading notes

Core claim

On the paper's own terms, the discovery is that the damage done by unlearning a skewed-class client is recoverable from the remaining clients' local data alone. By augmenting each remaining client's skewed-class samples in a learned low-dimensional feature space, then filtering the synthetic samples by a density factor computed against their local neighborhood, Imba-ULRc produces balanced, cleaner local datasets. Recovery training on these datasets drives the unlearning model back to high skewed-class accuracy: in the reported results it reaches 92.41 to 93.38 percent on MNIST, 89.07 to 89.77 percent on FMNIST, and 85.26 to 87.53 percent on USPS across skew levels alpha = 0.8, 0.85, and 0.9, outperforming the four baselines in most comparisons. The ablation shows the denoising step improves over SMOTE-only recovery in 16 of 18 cases.

Load-bearing premise

Each remaining client must hold at least one real sample of the skewed class, because the oversampling step interpolates from existing samples and the autoencoder needs that class during training.

Editorial extensions

If this is right

  • Federated unlearning no longer has to mean accepting a permanently biased model: after the leaving client's contribution is erased, the remaining clients can rebuild skewed-class accuracy through their own synthesized data.
  • The recovery works without access to the unlearning client's data or model, preserving the privacy boundary that motivates federated unlearning.
  • Because skewed-class accuracy stays roughly flat as skew alpha rises from 0.8 to 0.9, the method is claimed to be robust to how dominant the leaving client was in the skewed class.
  • The density-factor denoising step is a separable component: adding it to SMOTE-based recovery improved results in 16 of 18 experiments, so it can likely be attached to other oversampling-based recovery pipelines.

Reading between the lines

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

  • Editorial inference: the method implicitly assumes every remaining client has at least one real skewed-class sample; if some client has none, SMOTE's interpolation has no seed and the encoder has never seen that class. A practical extension would let clients borrow latent seeds from peers or re-weight aggregation for zero-shot clients.
  • Editorial inference: the median density-factor threshold is a heuristic with a testable alternative; replacing it with a quantile or a per-client calibrated threshold may trade recall of genuine boundary samples against noise removal, which the paper does not explore.
  • Editorial inference: since the synthetic data are generated locally and never leave each client, the approach may also dampen privacy leakage from raw data sharing, but the paper does not quantify how much information the synthetic samples or the trained autoencoder reveal.
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.

Referee Report

5 major / 6 minor

Summary. The paper addresses federated unlearning when the leaving client holds most of the data for one class (the skewed class). After unlearning, the remaining clients are assumed to have too little skewed-class data to recover the model. The proposed method, Imba-ULRc, has each remaining client train a local autoencoder, oversample the skewed class in the latent space with SMOTE (Eq. 4), denoise the generated samples using a density-factor threshold based on k-nearest neighbors (Section 3.3), and then perform recovery training on the unlearning model. The experimental section compares Imba-ULRc with four baselines on MNIST, FMNIST, and USPS under three skew levels alpha = 0.8, 0.85, 0.9. Table 1 reports that Imba-ULRc achieves the best skewed-class accuracy in all nine configurations and the best or near-best global accuracy in seven of nine, with ablations in Table 2 and a sensitivity analysis for k in Figure 5.

Significance. If the method holds in the regime it claims, it addresses a real gap in federated unlearning: recovering model quality after a skewed-label client leaves, without accessing that client's data. The empirical comparison uses standard public benchmarks, multiple baselines, and ablation studies, and the method is conceptually simple. However, the paper currently omits a key precondition (each remaining client must have enough real skewed-class samples), and the denoising parameter k is selected on the test sets used for final reporting. These issues narrow the validity of the central claim and weaken the evidence. The contribution is still potentially useful for the regime where every remaining client has at least a moderate number of skewed-class samples, but the manuscript must be revised to state and test this boundary. The paper does not provide code or machine-checked proofs; its evidence is purely empirical with three runs per setting.

major comments (5)
  1. [Section 3.2, Definition 1 and Eq. (4)] The method requires each remaining client to have at least one real sample of the skewed class to seed SMOTE interpolation in Eq. (4) and to train an autoencoder that can represent that class's manifold, but the paper never states this precondition. Definition 1 only states nu_C >> ni_J >> ni_C with no lower bound on ni_C, and the motivating example in Section 1 (small banks with much less high-net-worth data) naturally includes clients with zero such samples. The experimental protocol in Section 4.2 gives each remaining client (1-alpha)*N_C/4 samples, which is large for the tested alpha values, so the zero-sample case is never exercised. The abstract and introduction claim a general recovery method, but the method is undefined when ni_C = 0. Please state the required condition explicitly and either provide a fallback for zero-seed clients or scope the claims to the regime where every remaining client has at least a few skewed-class samples.
  2. [Section 5.3, Figure 5 and Table 1] The denoising parameter k is selected by varying k and measuring the skewed-class accuracy after recovery, apparently on the same test sets used to report the final results in Table 1. This is a form of test-set tuning, and it can selectively favor the proposed method when comparing against baselines that do not have access to the test set for parameter selection. Use a held-out validation split to select k and then report test accuracy for the chosen k, or report results across a range of k values and show that the main conclusions are insensitive to the choice.
  3. [Table 1 and Section 5.1] The caption of Table 1 says 'Balanced Accuracy of the Global Model', but the text repeatedly refers to this column as 'accuracy of the global model' or 'global model accuracy'. The reported values, such as 95.82 for MNIST at alpha=0.8, look like standard overall accuracy rather than balanced accuracy (macro-average recall). Since the test set may not be class-balanced, the term 'balanced accuracy' has a specific meaning that should be defined and used consistently. Clarify which metric is reported and ensure it is computed identically for all methods, or the comparison across methods may be misleading.
  4. [Section 3.3, Eqs. (5)-(6)] The density formulas have notation inconsistencies that prevent reproduction. Eq. (5) uses 'K + 1' in the denominator although the neighbor count is denoted by lowercase k elsewhere, and K is already used for the set of all classes in Definition 1. Eq. (6) has an inner sum over q' whose upper limit is written as 'Pk' in a way that is ambiguous. Since the denoising step is a central contribution, please rewrite these equations with clearly defined variables and consistent notation.
  5. [Section 5.1, Table 1] All comparisons are based on three runs with no statistical significance testing. The skewed-class accuracy gaps are large, but the global-accuracy differences are often small (e.g., MNIST alpha=0.8: Imba-ULRc 95.82 +/- 0.02 vs MOON 95.62 +/- 0.05; USPS alpha=0.8: Imba-ULRc 85.37 +/- 0.06 vs FLRS 85.50 +/- 0.09). With n=3, such differences may not be meaningful. Report individual run results or apply paired significance tests across runs and configurations to support the claim of consistent improvement.
minor comments (6)
  1. [Section 3.2, Eq. (3)] The reverse-order reconstruction loss is hard to follow because the notation mixes yp_i,c, y1_i, and dp_i,c without clear indexing. Explain the purpose of the reordering and define all indices explicitly.
  2. [Section 4.2] The text says 'five federated learning clients' and that Client 1 is unlearned, which leaves four remaining clients. State the number of remaining clients explicitly and report how many independent repetitions are used for each configuration.
  3. [Algorithm 1] The line 'where N dn_i denote the total amount of noised dataset Di_dn' uses 'noised' where 'denoised' is meant. This is a typo that could confuse readers.
  4. [Section 2.1, Eq. (1)] The distribution T is used in Eq. (1) without a definition. Clarify what T represents, e.g., the output distribution of the model.
  5. [Figure 5] The x-axis and y-axis labels are not described in the text or figure; state explicitly that the x-axis is the value of k and the y-axis is the skewed-class test accuracy.
  6. [References] Reference [11] is cited for SMOTE, but [11] is the DeepSMOTE paper; provide a reference to the original SMOTE paper by Chawla et al. and keep [11] only if DeepSMOTE is specifically used.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the recovery result is an empirical method comparison, self-contained against external benchmarks.

full rationale

The paper's central claim is empirical: after unlearning a skewed-class client, remaining clients oversample the skewed class in a learned latent space (Eq. 4), denoise generated samples using a density-factor threshold (Eqs. 5-6), and perform recovery training (Eqs. 7-8). None of these steps is derived from the result it is supposed to explain. The autoencoder and SMOTE are seeded by real remaining-client data, and the recovered model is evaluated on held-out test splits of MNIST, FMNIST, and USPS against four external baselines. There are no load-bearing self-citations, no uniqueness theorem imported from the authors, and no claimed prediction that is a fitted parameter under another name. The search for the denoising parameter k in Section 5.3 on the same benchmark test sets is an evaluation-protocol concern, but it is not a reduction of the recovery result to its inputs: the reported accuracies still depend on trained models and held-out labels. The zero-sample precondition for SMOTE in Definition 1 (nu_C >> ni_J >> ni_C leaves ni_C possibly zero) is a genuine scope limitation because Eq. 4 is undefined without same-class seed points; however, the experiments always allocate (1-alpha)*N_C/4 skewed-class samples to every remaining client, so this affects generality rather than circularity. The derivation chain is self-contained; no step is equivalent to its input by construction.

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

No new physical or conceptual entities are introduced; the generated data are synthetic examples, not a new modeling construct. The method's parameters and heuristics are listed above.

free parameters (3)
  • k (neighbor count for denoising) = 5
    Selected by sweeping k from 1 to 25 on the same test sets and choosing the value with the best skewed-class accuracy (Section 5.3, Figure 5); no held-out validation is reported.
  • SMOTE oversampling surplus = 2x needed
    The paper states that it generates twice as many samples as needed before denoising (Section 3.3); the factor of two is chosen by hand with no justification.
  • Density-factor threshold = median of density factors
    The removal threshold is defined as the median of the generated samples' density factors (Section 3.3), a data-dependent choice with no independent validation.
assumptions (4)
  • domain assumption Each remaining client possesses at least one real sample of the skewed class to seed SMOTE and train the autoencoder.
    Definition 1 assumes nu_C >> ni_J >> ni_C, implicitly ni_C > 0; Section 3.2 uses SMOTE nearest-neighbor interpolation, which is undefined for a client with zero skewed-class samples.
  • ad hoc to paper Interpolation in the autoencoder's latent space produces synthetic samples that lie on or near the true skewed-class manifold.
    Section 3.2 constructs x_i,g = x_i,C + Rand(0,1)*(x_i,N - x_i,C) and decodes it, assuming the decoder maps the interpolated latent point to a valid class-C sample; no direct evaluation of generated samples is provided.
  • ad hoc to paper The median-density-factor rule separates noise from valid generated samples.
    Section 3.3 assumes a right-skewed density-factor distribution and that points above the median are noise; this is a heuristic (LOF-like) with no theoretical guarantee.
  • domain assumption Retraining with FedAvg on enhanced local datasets restores the unlearning model's distribution.
    Section 3.4 assumes federated averaging of locally fine-tuned models converges to a model close to the pre-unlearning distribution; this is standard FL practice but not proven here.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Federated Unlearning Model Recovery in Data with Skewed Label Distributions." pith.science (2026). https://pith.science/paper/K7IUSTJR

@misc{pith2026241213466,
  author       = {Pith},
  title        = {Pith review of: Federated Unlearning Model Recovery in Data with Skewed Label Distributions},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/K7IUSTJR}},
  note         = {Machine review of arXiv:2412.13466}
}
read the original abstract

In federated learning, federated unlearning is a technique that provides clients with a rollback mechanism that allows them to withdraw their data contribution without training from scratch. However, existing research has not considered scenarios with skewed label distributions. Unfortunately, the unlearning of a client with skewed data usually results in biased models and makes it difficult to deliver high-quality service, complicating the recovery process. This paper proposes a recovery method of federated unlearning with skewed label distributions. Specifically, we first adopt a strategy that incorporates oversampling with deep learning to supplement the skewed class data for clients to perform recovery training, therefore enhancing the completeness of their local datasets. Afterward, a density-based denoising method is applied to remove noise from the generated data, further improving the quality of the remaining clients' datasets. Finally, all the remaining clients leverage the enhanced local datasets and engage in iterative training to effectively restore the performance of the unlearning model. Extensive evaluations on commonly used federated learning datasets with varying degrees of skewness show that our method outperforms baseline methods in restoring the performance of the unlearning model, particularly regarding accuracy on the skewed class.

Figures

Figures reproduced from arXiv: 2412.13466 by the authors.

Figure 1
Figure 1. Label distribution skew in the federated learning framework. The numbers in the boxes represent the amount [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. The overview of the proposed method. ➀ indicates the unlearning model’s performance bias caused by the skewed class’s presence in the skewed label distribution. ➁ and ➂ represent our proposed method for Addressing the Performance Bias. First, we train a data generation model composed of an encoder and decoder and then use SMOTE to oversample the skewed class of the remaining clients. ➃ and ➄ represent our proposed D… view at source ↗
Figure 3
Figure 3. The density-based method for removing noise from generated data. The orange dots represent skewed class, [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: Details of the recovery training. the local recovery model ω t i is obtained: ω t i = ωu − η∇L(ωu, Di dn), (7) where η represents the learning rate. In [PITH_FULL_IMAGE:figures/full_fig_p008_4.png]
Figure 5
Figure 5. Figure 5: Under three data skew settings, the accuracy of the model on skewed class after client-generated data denoising [PITH_FULL_IMAGE:figures/full_fig_p012_5.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

42 extracted references · 29 canonical work pages

  1. [1]

    Communication- efficient learning of deep networks from decentralized data

    Brendan McMahan, Eider Moore, Daniel Ramage, Seth Hampson, and Blaise Aguera y Arcas. Communication- efficient learning of deep networks from decentralized data. In Artificial intelligence and statistics , pages 1273–1282. PMLR, 2017

    2017

  2. [2]

    A survey on federated learning.Knowledge- Based Systems, 216:106775, 2021

    Chen Zhang, Yu Xie, Hang Bai, Bin Yu, Weihong Li, and Yuan Gao. A survey on federated learning.Knowledge- Based Systems, 216:106775, 2021

    work page 2021

  3. [3]

    A survey on federated unlearning: Challenges, methods, and future directions

    Ziyao Liu, Yu Jiang, Jiyuan Shen, Minyi Peng, Kwok-Yan Lam, Xingliang Yuan, and Xiaoning Liu. A survey on federated unlearning: Challenges, methods, and future directions. ACM Computing Surveys, 2023

    work page 2023

  4. [4]

    Federated Unlearning: A Survey on Methods, Design Guidelines, and Evaluation Metrics

    Nicolò Romandini, Alessio Mora, Carlo Mazzocca, Rebecca Montanari, and Paolo Bellavista. Federated unlearning: A survey on methods, design guidelines, and evaluation metrics. arXiv preprint arXiv:2401.05146, 2024

    work page Pith review arXiv 2024

  5. [5]

    SoK: Challenges and opportunities in federated unlearning

    Hyejun Jeong, Shiqing Ma, and Amir Houmansadr. SoK: Challenges and opportunities in federated unlearning. arXiv preprint arXiv:2403.02437, 2024

    arXiv 2024

  6. [6]

    Federated unlearning and its privacy threats

    Fei Wang, Baochun Li, and Bo Li. Federated unlearning and its privacy threats. IEEE Network, 38(2):294–300, 2023

    work page 2023

  7. [7]

    Modeling global distribution for federated learning with label distribution skew

    Tao Sheng, Chengchao Shen, Yuan Liu, Yeyu Ou, Zhe Qu, Yixiong Liang, and Jianxin Wang. Modeling global distribution for federated learning with label distribution skew. Pattern Recognition, 143:109724, 2023

    work page 2023

  8. [8]

    Federated learning with label distribution skew via logits calibration

    Jie Zhang, Zhiqi Li, Bo Li, Jianghe Xu, Shuang Wu, Shouhong Ding, and Chao Wu. Federated learning with label distribution skew via logits calibration. In International Conference on Machine Learning, pages 26311–26329. PMLR, 2022

    work page 2022

Show all 42 references
  1. [9]

    Fedrs: Federated learning with restricted softmax for label distribution non-iid data

    Xin-Chun Li and De-Chuan Zhan. Fedrs: Federated learning with restricted softmax for label distribution non-iid data. In Proceedings of the 27th ACM SIGKDD Conference on Knowledge Discovery & Data Mining , pages 995–1005, 2021

    2021

  2. [10]

    Quickdrop: Efficient federated unlearning by integrated dataset distillation

    Akash Dhasade, Yaohong Ding, Song Guo, Anne-marie Kermarrec, Martijn De V os, and Leijie Wu. Quickdrop: Efficient federated unlearning by integrated dataset distillation. arXiv preprint arXiv:2311.15603, 2023

    2023 arXiv

  3. [11]

    Deepsmote: Fusing deep learning and smote for imbalanced data

    Damien Dablain, Bartosz Krawczyk, and Nitesh V Chawla. Deepsmote: Fusing deep learning and smote for imbalanced data. IEEE Transactions on Neural Networks and Learning Systems, 34(9):6390–6404, 2022

    2022

  4. [12]

    The eu general data protection regulation (gdpr).A Practical Guide, 1st Ed., Cham: Springer International Publishing, 10(3152676):10–5555, 2017

    Paul V oigt and Axel V on dem Bussche. The eu general data protection regulation (gdpr).A Practical Guide, 1st Ed., Cham: Springer International Publishing, 10(3152676):10–5555, 2017

    2017

  5. [13]

    The california consumer privacy act of 2018

    KJ Mathews and CM Bowman. The california consumer privacy act of 2018

    2018

  6. [14]

    The right to be forgotten in federated learning: An efficient realization with rapid retraining

    Yi Liu, Lei Xu, Xingliang Yuan, Cong Wang, and Bo Li. The right to be forgotten in federated learning: An efficient realization with rapid retraining. In IEEE INFOCOM 2022-IEEE Conference on Computer Communications, pages 1749–1758. IEEE, 2022. 13 A PREPRINT - D ECEMBER 23, 2024

    2022

  7. [15]

    Federated unlearning: How to efficiently erase a client in fl? arXiv preprint arXiv:2207.05521, 2022

    Anisa Halimi, Swanand Kadhe, Ambrish Rawat, and Nathalie Baracaldo. Federated unlearning: How to efficiently erase a client in fl? arXiv preprint arXiv:2207.05521, 2022

    2022 arXiv

  8. [16]

    Federated unlearning: Guarantee the right of clients to forget

    Leijie Wu, Song Guo, Junxiao Wang, Zicong Hong, Jie Zhang, and Yaohong Ding. Federated unlearning: Guarantee the right of clients to forget. IEEE Network, 36(5):129–135, 2022

    2022

  9. [17]

    Learn to forget: User-level memorization elimination in federated learning

    Yang Liu, Zhuo Ma, Ximeng Liu, and Jianfeng Ma. Learn to forget: User-level memorization elimination in federated learning. arXiv preprint arXiv:2003.10933, 1, 2020

    2003 arXiv

  10. [18]

    Exact-fun: An exact and efficient federated unlearning approach

    Zuobin Xiong, Wei Li, Yingshu Li, and Zhipeng Cai. Exact-fun: An exact and efficient federated unlearning approach. In 2023 IEEE International Conference on Data Mining (ICDM), pages 1439–1444, 2023

    2023

  11. [19]

    Forget-svgd: Particle-based bayesian federated unlearning

    Jinu Gong, Joonhyuk Kang, Osvaldo Simeone, and Rahif Kassab. Forget-svgd: Particle-based bayesian federated unlearning. In 2022 IEEE Data Science and Learning Workshop (DSLW), pages 1–6, 2022

    2022

  12. [20]

    Fedrecovery: Differentially private machine unlearning for federated learning frameworks

    Lefeng Zhang, Tianqing Zhu, Haibin Zhang, Ping Xiong, and Wanlei Zhou. Fedrecovery: Differentially private machine unlearning for federated learning frameworks. IEEE Transactions on Information Forensics and Security, 2023

    2023

  13. [21]

    Verifi: Towards verifiable federated unlearning

    Xiangshan Gao, Xingjun Ma, Jingyi Wang, Youcheng Sun, Bo Li, Shouling Ji, Peng Cheng, and Jiming Chen. Verifi: Towards verifiable federated unlearning. IEEE Transactions on Dependable and Secure Computing, 2024

    2024

  14. [22]

    Sifu: Sequential informed federated unlearning for efficient and provable client unlearning in federated optimization

    Yann Fraboni, Martin Van Waerebeke, Kevin Scaman, Richard Vidal, Laetitia Kameni, and Marco Lorenzi. Sifu: Sequential informed federated unlearning for efficient and provable client unlearning in federated optimization. In International Conference on Artificial Intelligence an...

    2024

  15. [23]

    Federated unlearning via class-discriminative pruning

    Junxiao Wang, Song Guo, Xin Xie, and Heng Qi. Federated unlearning via class-discriminative pruning. In Proceedings of the ACM Web Conference 2022, pages 622–632, 2022

    2022

  16. [24]

    A state-of-the-art survey on solving non-iid data in federated learning

    Xiaodong Ma, Jia Zhu, Zhihao Lin, Shanxuan Chen, and Yangjie Qin. A state-of-the-art survey on solving non-iid data in federated learning. Future Generation Computer Systems, 135:244–258, 2022

    2022

  17. [25]

    Federated learning on non-iid data silos: An experimental study

    Qinbin Li, Yiqun Diao, Quan Chen, and Bingsheng He. Federated learning on non-iid data silos: An experimental study. In 2022 IEEE 38th International Conference on Data Engineering (ICDE), pages 965–978. IEEE, 2022

    2022

  18. [26]

    Federated learning with non-iid data: A survey

    Zili Lu, Heng Pan, Yueyue Dai, Xueming Si, and Yan Zhang. Federated learning with non-iid data: A survey. IEEE Internet of Things Journal, 2024

    2024

  19. [27]

    Robust and communication- efficient federated learning from non-iid data

    Felix Sattler, Simon Wiedemann, Klaus-Robert Müller, and Wojciech Samek. Robust and communication- efficient federated learning from non-iid data. IEEE Transactions on Neural Networks and Learning Systems, 31(9):3400–3413, 2019

    2019

  20. [28]

    Scaffold: Stochastic controlled averaging for federated learning

    Sai Praneeth Karimireddy, Satyen Kale, Mehryar Mohri, Sashank Reddi, Sebastian Stich, and Ananda Theertha Suresh. Scaffold: Stochastic controlled averaging for federated learning. In International Conference on Machine Learning, pages 5132–5143. PMLR, 2020

    2020

  21. [29]

    Model-contrastive federated learning

    Qinbin Li, Bingsheng He, and Dawn Song. Model-contrastive federated learning. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 10713–10722, 2021

    2021

  22. [30]

    Data distribution-aware online client selection algorithm for federated learning in heterogeneous networks

    Jaewook Lee, Haneul Ko, Sangwon Seo, and Sangheon Pack. Data distribution-aware online client selection algorithm for federated learning in heterogeneous networks. IEEE Transactions on Vehicular Technology , 72(1):1127–1136, 2022

    2022

  23. [31]

    Adaptive federated learning on non-iid data with resource constraint

    Jie Zhang, Song Guo, Zhihao Qu, Deze Zeng, Yufeng Zhan, Qifeng Liu, and Rajendra Akerkar. Adaptive federated learning on non-iid data with resource constraint. IEEE Transactions on Computers, 71(7):1655–1667, 2021

    2021

  24. [32]

    Federated learning enabled hotel customer classification towards imbalanced data

    Tao Liu, Shouqiang Chen, Meng Wu, and Miao Yu. Federated learning enabled hotel customer classification towards imbalanced data. Applied Soft Computing, 166:112028, 2024

    2024

  25. [33]

    Federated learning with non-iid data

    Yue Zhao, Meng Li, Liangzhen Lai, Naveen Suda, Damon Civin, and Vikas Chandra. Federated learning with non-iid data. arXiv preprint arXiv:1806.00582, 2018

    2018 arXiv

  26. [34]

    Distillation-based semi-supervised federated learning for communication-efficient collaborative training with non-iid private data

    Sohei Itahara, Takayuki Nishio, Yusuke Koda, Masahiro Morikura, and Koji Yamamoto. Distillation-based semi-supervised federated learning for communication-efficient collaborative training with non-iid private data. IEEE Transactions on Mobile Computing, 22(1):191–205, 2023

    2023

  27. [35]

    Towards asynchronous federated learning based threat detection: A dc-adam approach

    Pu Tian, Zheyi Chen, Wei Yu, and Weixian Liao. Towards asynchronous federated learning based threat detection: A dc-adam approach. Computers & Security, 108:102344, 2021

    2021

  28. [36]

    Dynamic margin for federated learning with imbalanced data

    Xinyu Ran, Liang Ge, and Linlin Zhong. Dynamic margin for federated learning with imbalanced data. In 2021 International Joint Conference on Neural Networks (IJCNN), pages 1–8. IEEE, 2021. 14 A PREPRINT - D ECEMBER 23, 2024

    2021

  29. [37]

    Lof: identifying density-based local outliers

    Markus M Breunig, Hans-Peter Kriegel, Raymond T Ng, and Jörg Sander. Lof: identifying density-based local outliers. In Proceedings of the 2000 ACM SIGMOD International Conference on Management of Data, pages 93–104, 2000

    2000

  30. [38]

    Anomaly detection: A survey

    Varun Chandola, Arindam Banerjee, and Vipin Kumar. Anomaly detection: A survey. ACM computing surveys (CSUR), 41(3):1–58, 2009

    2009

  31. [39]

    Loci: Fast outlier detection using the local correlation integral

    Spiros Papadimitriou, Hiroyuki Kitagawa, Phillip B Gibbons, and Christos Faloutsos. Loci: Fast outlier detection using the local correlation integral. In Proceedings 19th International Conference on Data Engineering (Cat. No. 03CH37405), pages 315–326. IEEE, 2003

    2003

  32. [40]

    The mnist database of handwritten digit images for machine learning research [best of the web]

    Li Deng. The mnist database of handwritten digit images for machine learning research [best of the web]. IEEE Signal Processing Magazine, 29(6):141–142, 2012

    2012

  33. [41]

    Fashion-mnist: a novel image dataset for benchmarking machine learning algorithms

    Han Xiao, Kashif Rasul, and Roland V ollgraf. Fashion-mnist: a novel image dataset for benchmarking machine learning algorithms. arXiv preprint arXiv:1708.07747, 2017

    2017 arXiv

  34. [42]

    Jonathan J. Hull. A database for handwritten text recognition research. IEEE Transactions on Pattern Analysis and Machine Intelligence, 16(5):550–554, 1994. 15

    1994

Pith tools

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