Pith. sign in

REVIEW 3 major objections 5 minor 51 references

Federated Learning in Mobile Networks: A Comprehensive Case Study on Traffic Forecasting

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 Federated learning beats local and centralized training for mobile traffic forecasting on real base-station data, the paper claims, and does so while saving communication and preserving privacy.

desk verdict Useful applied FL study with a new Barcelona dataset and solid ablations, but the sustainability table doesn't reproduce from the paper's own equations — fix that before trusting the main claim. read the letter →

arxiv 2412.04081 v1 pith:LSNW4TUJ submitted 2024-12-05 cs.LG cs.AI

Nikolaos Pavlidis , Vasileios Perifanis , Selim F. Yilmaz , Francesc Wilhelmi , Marco Miozzo , Pavlos S. Efraimidis , Remous-Aris Koutsiamanis , Pavol Mulinka
show 1 more author
Paolo Dini
This is my paper · ORCID
classification cs.LGcs.AI
keywords federatedlearningmobiletrafficforecastinglongshort-termmemory(LSTM)sustainabilityindicatornon-IIDdataclientselectionexogenousenrichment5G/6Gnetworks
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 argues that federated learning—where base stations exchange model updates rather than raw measurements—is the best among three training regimes for cellular traffic forecasting on a real seven-site dataset from Barcelona. The authors compare a federated LSTM model with each base station training alone and with all data pooled at one central server, controlling for the number of dataset accesses. Federated learning achieves the lowest average prediction error (NRMSE 1.385) and the best composite sustainability score (S = 30.24), which balances error, training energy, transmitted data size, and inference energy. The result matters because mobile traffic forecasts feed resource allocation and anomaly detection in 5G and future 6G networks, where a privacy-preserving option that does not sacrifice accuracy or energy efficiency is directly useful to operators.

What carries the argument

The load-bearing mechanism is the federated averaging loop around an LSTM predictor: selected base stations download the global model, train locally for a few epochs, send back weight updates, and the server combines them as a dataset-size-weighted average. The paper's main quantitative instrument is the composite sustainability indicator $S = S_{\mathrm{Tr}} \times S_{\mathrm{Inf}}$, with $S_{\mathrm{Tr}} = (1 + E_{\mathrm{Val}})^\alpha(1 + C_{\mathrm{Tr}})^\beta(1 + D_S)^\gamma$ and $S_{\mathrm{Inf}} = E_{\mathrm{Test}}^{\alpha'} C_{\mathrm{Inf}}^{\beta'}$, where the hand-set weights are $\alpha = \beta = \gamma = 1/3$ and $\alpha' = \beta' = 0.5$; the tiny transmitted model size is what gives FL its large advantage. The LSTM itself (stacked recurrent layers followed by a ReLU feed-forward head and a linear projection) is adopted because an earlier study found it the best accuracy-to-efficiency balance among RNN, CNN, GRU, and transformer alternatives.

What would settle it

Recompute the sustainability score S from the paper's own Table IV numbers with the energy exponent raised from 1/3 to, say, 0.6; if individual training's lower energy (13.03 Wh versus 14 Wh) then beats federated learning's score, the conclusion that FL is the most sustainable setting fails.

Watch

Extended reading notes

Core claim

The paper's central claim is that on the Barcelona PDCCH dataset, federated collaboration dominates both alternatives: it reaches an average NRMSE of 1.385, below centralized training's 1.434 and individual training's 1.92, while consuming about 14 Wh of training energy, roughly 8% more than individual training and 10% less than centralized training. On the sustainability indicator S, federated learning scores 30.24 against 71.67 for individual training and 132.03 for centralized training, because it transmits only 217 kB of model updates per round instead of the full 16,531 kB dataset. Ablation studies reinforce the central claim: FedAvg matches more complex aggregators, isolation-forest capping of outliers improves accuracy, local fine-tuning lowers the federated error by about 10.9%, and enriched exogenous features reduce the average error by 6.55% though with mixed per-site results.

Load-bearing premise

The claim that federated learning is the most sustainable setting rests on the paper's hand-picked weights in the sustainability formula, and those weights are never varied, even though individual training actually uses less energy.

Editorial extensions

If this is right

  • Operators can obtain near-centralized accuracy without moving raw traffic data off base stations, reducing privacy exposure and data-transfer cost.
  • Under the paper's matched-access protocol, the privacy-preserving option costs only about 8% more training energy than per-site training, so privacy does not carry a large energy penalty.
  • Cutting participation from 7/7 to 5/7 sites per round degrades average error by only 4–5%, so client selection is a practical lever for lowering per-round cost.
  • Local fine-tuning after the federated round improves per-site error by roughly 10.9%, making FL workable across non-IID site traffic.
  • Exogenous data (holidays, events, calendar features) lowers average federated error by 6.55%, but gains are uneven across sites with anomalous events.

Reading between the lines

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

  • If the sustainability weights were shifted toward energy consumption, individual training's lower measured cost (13.03 Wh) could overtake FL's 14 Wh and break the claimed ranking; the missing sensitivity analysis is the natural next check.
  • Centralized error decays more slowly at long horizons (from step 8 onward), suggesting that pooled data is most valuable for distant forecasts; a hybrid that federates short horizons and centralizes long ones is a testable extension.
  • Removing site E2 hurt site S1 more than removing S1 itself, and the paper links this to low KL divergence between those datasets; this turns the diagnostic into a concrete candidate rule for data-similarity-aware client selection.
  • The mixed effect of exogenous enrichment implies feature selection should be per-site rather than global, since sites with one-off events did not benefit from the automatically added features.
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

3 major / 5 minor

Summary. The paper presents a case study of federated learning (FL) for cellular traffic forecasting on real LTE data from seven Barcelona base-station sites. It compares three training settings—Individual, Centralized, and Federated—using an LSTM model, reporting prediction error (NRMSE/MAE) and a sustainability indicator S that combines error, training energy, transmitted data size, and inference energy. The paper also reports ablations on outlier handling, model aggregation, client selection, personalized fine-tuning, and exogenous data enrichment. The central claims are that the federated setting achieves the lowest average prediction error (NRMSE 1.385) and the best accuracy-energy-communication trade-off (S = 30.24), making FL a privacy-preserving and sustainable alternative for traffic forecasting.

Significance. If the claims held, this would be a useful empirical benchmark for FL in mobile traffic forecasting, with the strengths of using a real operator dataset, running 10 seeds, ensuring equal dataset accesses across settings, and providing source code. The ablation studies cover relevant practical aspects (outliers, aggregation, client selection, personalization, exogenous data) that are often missing in FL case studies. However, the main sustainability conclusion rests on a non-reproducible computation in Table IV, and the headline accuracy advantage over centralized learning is small and not supported by any statistical test. The case-study nature and the hand-chosen sustainability weights further limit the generality of the conclusions.

major comments (3)
  1. [Section III-B, Eqs. (6)-(8) and Table IV] The reported STr, SInf, and S values in Table IV cannot be reproduced from the paper's own formulas and the input values given in the same table. Using alpha=beta=gamma=1/3 and alpha'=beta'=0.5, the Individual setting yields STr = (1+1.92)^(1/3)*(1+13.03)^(1/3)*(1+0)^(1/3) ≈ 3.45 and SInf = sqrt(1.92)*sqrt(0.048) ≈ 0.30, giving S ≈ 1.05, not the reported STr=40.98, SInf=1.75, S=71.68. For the Federated setting, the same formulas give STr ≈ 19.84 and SInf ≈ 0.20, not 19.27 and 1.57. No alternative normalization, error metric, or rounding convention is stated. Since the text uses Table IV to state that 'FL's superiority over other settings is proven', the sustainability claim is unsupported as written. Please correct the table, state the exact definitions of EVal and ETest and any normalization used, or provide the script that computes S.
  2. [Section V-A, Table III] The headline accuracy result is Federated NRMSE 1.385 ± 0.044 versus Centralized 1.434 ± 0.003 and Individual 1.92 ± 0.0235. The difference between Federated and Centralized is only 0.049, which is smaller than the reported standard deviation of the Federated mean, and no significance test (e.g., paired t-test over seeds or per-site Wilcoxon test) is reported. The sentence that the federated setting 'achieves the lowest average error of 1.385' therefore overstates the evidence; a confidence interval or significance test is needed to support the claimed superiority.
  3. [Section III-B and Table II] The sustainability indicator S is a weighted product with hand-chosen exponents alpha=beta=gamma=1/3 and alpha'=beta'=0.5. The paper provides no sensitivity analysis for these weights. Because Table IV shows that the Individual setting consumes less training energy (13.03 Wh) than Federated (14.06 Wh), a weight profile that favors energy over accuracy could change the ranking among settings. The claim that 'FL gets the lowest S value ... thus achieving the best trade-off' needs robustness checks; please report how S changes under at least a few alternative weight choices.
minor comments (5)
  1. [Introduction and Section IV-A] The introduction says the data were collected at 'five base stations', but the dataset section lists seven sites/splits (LCCN, PS, EB, S1, S2, E1, E2); please align the descriptions.
  2. [Section III-B] The variables EVal and ETest in Eqs. (7) and (8) are not explicitly mapped to a specific error metric (NRMSE or MAE); please define them precisely in the text.
  3. [Tables III and IV] The total training energy is reported as 13 Wh in Table III and 13.03 Wh in Table IV, and as 14 Wh versus 14.06 Wh for Federated; use consistent significant figures across tables.
  4. [Section I, contributions] The first contribution bullet says 'we introduce ... a novel indicator', but Section III-B says the indicator S was 'introduced in [28]'; please rephrase to 'adapted' or 'applied' to avoid inconsistency.
  5. [Section V-B] Reference [48] is cited as 'Tzu et al.' in the aggregation subsection; the actual first author is Hsu et al., so the citation text should be corrected.

Circularity Check

0 steps flagged · score 2.0 of 10

No derivation-by-construction in the paper; the main accuracy results are empirical and self-contained, but the sustainability conclusion depends on a self-authored indicator with hand-assigned weights and the reported S values do not reproduce from the stated equations.

full rationale

The paper does not exhibit a circular derivation: no equation is defined in terms of the target conclusion, and no fitted parameter is relabeled as a prediction. The central accuracy comparison (Table III) is an empirical result from training LSTM models under three settings, and the claim that the federated setting achieves the lowest average NRMSE (1.385) is independent of any self-citation. The sustainability indicator S is taken from the authors' prior work [28], and the LSTM model choice is justified by the authors' earlier study [13]; these are self-citations with overlapping authors, but the S formula is stated in Eqs. (6)-(8), the weights are explicitly listed in Table II, and the rankings are not forced by construction, since a different weight profile or different measured values could change the outcome. A notable correctness concern is that the STr, SInf, and S values in Table IV cannot be reproduced from the paper's own Eqs. (6)-(8) with the stated inputs and weights (e.g., for Individual, (1+1.92)^(1/3)*(1+13.03)^(1/3)*(1+0)^(1/3) = 3.45, not 40.98). This undermines the support for the sustainability conclusion, but it is a reproducibility and internal-consistency flaw, not circularity. Therefore, the circularity score is low: no significant circularity, though the headline sustainability result rests on an unverified and internally inconsistent metric computation.

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

The paper introduces no new entities. It does rely on a hand-weighted sustainability metric from the authors' earlier work, on the representativeness of a seven-client Barcelona dataset, and on the transferability of laptop energy measurements to network edge hardware. LSTM hyperparameters needed for reproduction are not reported.

free parameters (3)
  • Sustainability indicator weights (alpha, beta, gamma) = 0.33, 0.33, 0.33
    Used in Eq. (7) to trade off validation error, training energy, and transmitted data size. The ranking of Individual/Centralized/Federated on S depends on these values, yet no sensitivity analysis is provided.
  • Sustainability indicator weights (alpha', beta') = 0.5, 0.5
    Used in Eq. (8) for inference sustainability; chosen without justification.
  • LSTM architecture hyperparameters = not reported
    Number of layers, hidden units, learning rate, batch size, and sequence length are absent from Table II; the accuracy results cannot be reproduced without these choices.
assumptions (3)
  • domain assumption The Barcelona PDCCH dataset from seven client splits is representative of mobile network traffic for drawing general conclusions about federated learning.
    Section IV uses data from five locations in one city with different collection periods; the abstract extends conclusions beyond this dataset.
  • domain assumption The sustainability indicator S from the authors' prior work [28] is a valid measure of environmental trade-offs.
    Section III-B adopts S without independent justification; the claim of FL's sustainability depends on this metric.
  • domain assumption Energy consumption measured with CarbonTracker on an AMD Ryzen 5 laptop represents energy use in real edge/base-station deployments.
    Section V states all experiments ran on a Windows 11 workstation; this hardware is not a base station.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Federated Learning in Mobile Networks: A Comprehensive Case Study on Traffic Forecasting." pith.science (2026). https://pith.science/paper/LSNW4TUJ

@misc{pith2026241204081,
  author       = {Pith},
  title        = {Pith review of: Federated Learning in Mobile Networks: A Comprehensive Case Study on Traffic Forecasting},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/LSNW4TUJ}},
  note         = {Machine review of arXiv:2412.04081}
}
read the original abstract

The increasing demand for efficient resource allocation in mobile networks has catalyzed the exploration of innovative solutions that could enhance the task of real-time cellular traffic prediction. Under these circumstances, federated learning (FL) stands out as a distributed and privacy-preserving solution to foster collaboration among different sites, thus enabling responsive near-the-edge solutions. In this paper, we comprehensively study the potential benefits of FL in telecommunications through a case study on federated traffic forecasting using real-world data from base stations (BSs) in Barcelona (Spain). Our study encompasses relevant aspects within the federated experience, including model aggregation techniques, outlier management, the impact of individual clients, personalized learning, and the integration of exogenous sources of data. The performed evaluation is based on both prediction accuracy and sustainability, thus showcasing the environmental impact of employed FL algorithms in various settings. The findings from our study highlight FL as a promising and robust solution for mobile traffic prediction, emphasizing its twin merits as a privacy-conscious and environmentally sustainable approach, while also demonstrating its capability to overcome data heterogeneity and ensure high-quality predictions, marking a significant stride towards its integration in mobile traffic management systems.

Figures

Figures reproduced from arXiv: 2412.04081 by the authors.

Figure 1
Figure 1. Overview of federated learning traffic prediction involving multiple [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Average KL-divergence between datasets from different BSs. [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 3
Figure 3. Special events that affect the RNTI count in LCCN, EB, and S1. [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figures from the paper (4 more)
Figure 4
Figure 4. Figure 4: Mean and standard deviation of the MAE obtained by each learning [PITH_FULL_IMAGE:figures/full_fig_p007_4.png]
Figure 6
Figure 6. Figure 6: Mean and standard deviation of the MAE achieved by each model [PITH_FULL_IMAGE:figures/full_fig_p008_6.png]
Figure 7
Figure 7. Figure 7: Importance heatmap of original and exogenous features (highlighted [PITH_FULL_IMAGE:figures/full_fig_p010_7.png]
Figure 8
Figure 8. Figure 8: Test MAE across various BS with the presence or absence of [PITH_FULL_IMAGE:figures/full_fig_p010_8.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

51 extracted references · 40 canonical work pages

  1. [1]

    Characterizing and modeling internet traffic dy- namics of cellular devices,

    M. Z. Shafiq et al. , “Characterizing and modeling internet traffic dy- namics of cellular devices,”ACM SIGMETRICS Performance Evaluation Review, vol. 39, no. 1, pp. 265–276, 2011

    work page 2011

  2. [2]

    Machine learning for networking: Workflow, advances and opportunities,

    M. Wang et al., “Machine learning for networking: Workflow, advances and opportunities,” Ieee Network, vol. 32, no. 2, pp. 92–99, 2017. 11

    work page 2017

  3. [3]

    Building a digital twin for network optimiza- tion using graph neural networks,

    M. Ferriol-Galm ´es et al., “Building a digital twin for network optimiza- tion using graph neural networks,” Computer Networks , vol. 217, p. 109329, 2022

    work page 2022

  4. [4]

    Traffic flow prediction with big data: a deep learning approach,

    Y . Lv et al. , “Traffic flow prediction with big data: a deep learning approach,” IEEE Transactions on Intelligent Transportation Systems , vol. 16, no. 2, pp. 865–873, 2014

    work page 2014

  5. [5]

    Deep learning on traffic prediction: Methods, analysis and future directions,

    X. Yin et al. , “Deep learning on traffic prediction: Methods, analysis and future directions,” IEEE Transactions on Intelligent Transportation Systems, 2021

    work page 2021

  6. [6]

    Long-term mobile traffic forecasting using deep spatio-temporal neural networks,

    C. Zhang and P. Patras, “Long-term mobile traffic forecasting using deep spatio-temporal neural networks,” in Proceedings of the Eighteenth ACM International Symposium on Mobile Ad Hoc Networking and Computing, 2018, pp. 231–240

    work page 2018

  7. [7]

    How much data is needed to train a medical image deep learning system to achieve necessary high accuracy?

    J. Cho et al., “How much data is needed to train a medical image deep learning system to achieve necessary high accuracy?” arXiv preprint arXiv:1511.06348, 2015

    arXiv 2015

  8. [8]

    Mobile Traffic Forecasting for Green 5G Networks,

    N. Piovesan et al., “Mobile Traffic Forecasting for Green 5G Networks,” in 2021 IEEE Global Communications Conference (GLOBECOM) . IEEE, 2021, pp. 1–6

    work page 2021

Show all 51 references
  1. [9]

    Forecasting mobile traffic to achieve greener 5G net- works: When machine learning is key,

    N. Piovesan et al., “Forecasting mobile traffic to achieve greener 5G net- works: When machine learning is key,” in 2021 IEEE 22nd International Workshop on Signal Processing Advances in Wireless Communications (SPAWC). IEEE, 2021, pp. 276–280

    2021

  2. [10]

    Estimating energy consumption of cloud, fog, and edge computing infrastructures,

    E. Ahvar, A.-C. Orgerie, and A. Lebre, “Estimating energy consumption of cloud, fog, and edge computing infrastructures,” IEEE Transactions on Sustainable Computing , vol. 7, no. 2, pp. 277–288, 2022

    2022

  3. [11]

    Communication-efficient learning of deep networks from decentralized data,

    B. McMahan et al., “Communication-efficient learning of deep networks from decentralized data,” in Artificial intelligence and statistics. PMLR, 2017, pp. 1273–1282

    2017

  4. [12]

    Federated learning: Strategies for improving com- munication efficiency,

    J. Kone ˇcn`y et al. , “Federated learning: Strategies for improving com- munication efficiency,” arXiv preprint arXiv:1610.05492 , 2016

    2016 arXiv

  5. [13]

    Federated learning for 5G base station traffic forecasting,

    V . Perifanis et al. , “Federated learning for 5G base station traffic forecasting,” Computer Networks, vol. 235, p. 109950, 2023

    2023

  6. [14]

    Mobile traffic classification through physical control channel fingerprinting: a deep learning approach,

    H. D. Trinh et al., “Mobile traffic classification through physical control channel fingerprinting: a deep learning approach,” IEEE Transactions on Network and Service Management, vol. 18, no. 2, pp. 1946–1961, 2020

    1946

  7. [15]

    Time series analysis: Forecasting and control San Francisco,

    G. E. Box and G. M. Jenkins, “Time series analysis: Forecasting and control San Francisco,” Calif: Holden-Day, 1976

    1976

  8. [16]

    Neural network models for time series forecasts,

    T. Hill, M. O’Connor, and W. Remus, “Neural network models for time series forecasts,” Management science , vol. 42, no. 7, pp. 1082–1092, 1996

    1996

  9. [17]

    Recurrent neural network based language model,

    T. Mikolov et al., “Recurrent neural network based language model,” in Interspeech, vol. 2, no. 3. Makuhari, 2010, pp. 1045–1048

    2010

  10. [18]

    Speech recognition with deep recurrent neural networks,

    A. Graves, A.-r. Mohamed, and G. Hinton, “Speech recognition with deep recurrent neural networks,” in 2013 IEEE international conference on acoustics, speech and signal processing. Ieee, 2013, pp. 6645–6649

    2013

  11. [19]

    Deepfake video detection using recurrent neural networks,

    D. G ¨uera and E. J. Delp, “Deepfake video detection using recurrent neural networks,” in 2018 15th IEEE international conference on advanced video and signal based surveillance (AVSS) . IEEE, 2018, pp. 1–6

    2018

  12. [20]

    Mobile traffic prediction from raw data using LSTM networks,

    H. D. Trinh, L. Giupponi, and P. Dini, “Mobile traffic prediction from raw data using LSTM networks,” in 2018 IEEE 29th annual interna- tional symposium on personal, indoor and mobile radio communications (PIMRC). IEEE, 2018, pp. 1827–1832

    2018

  13. [21]

    Urban anomaly detection by processing mobile traffic traces with LSTM neural networks,

    H. D. Trinh, L. Giupponi, and P. Dini, “Urban anomaly detection by processing mobile traffic traces with LSTM neural networks,” in 2019 16th Annual IEEE International Conference on Sensing, Communica- tion, and Networking (SECON) . IEEE, 2019, pp. 1–8

    2019

  14. [22]

    Spatio-temporal wireless traffic prediction with recurrent neural network,

    C. Qiu et al., “Spatio-temporal wireless traffic prediction with recurrent neural network,” IEEE Wireless Communications Letters , vol. 7, no. 4, pp. 554–557, 2018

    2018

  15. [23]

    Recent advances in convolutional neural networks,

    J. Gu et al., “Recent advances in convolutional neural networks,” Pattern recognition, vol. 77, pp. 354–377, 2018

    2018

  16. [24]

    Network traffic prediction based on diffusion convolutional recurrent neural networks,

    D. Andreoletti et al. , “Network traffic prediction based on diffusion convolutional recurrent neural networks,” in IEEE INFOCOM 2019- IEEE Conference on Computer Communications Workshops (INFOCOM WKSHPS). IEEE, 2019, pp. 246–251

    2019

  17. [25]

    Deep residual learning for image recognition,

    K. He et al., “Deep residual learning for image recognition,” in Proceed- ings of the IEEE conference on computer vision and pattern recognition, 2016, pp. 770–778

    2016

  18. [26]

    Very deep convolutional networks for large-scale image recognition,

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

    2014 arXiv

  19. [27]

    A multivariate-time-series-prediction-based adaptive data transmission period control algorithm for iot networks,

    J. Han et al., “A multivariate-time-series-prediction-based adaptive data transmission period control algorithm for iot networks,” IEEE Internet of Things Journal , vol. 9, no. 1, pp. 419–436, 2021

    2021

  20. [28]

    Towards energy-aware federated traffic prediction for cellular networks,

    V . Perifanis et al. , “Towards energy-aware federated traffic prediction for cellular networks,” in 2023 Eighth International Conference on Fog and Mobile Edge Computing (FMEC) . IEEE, 2023, pp. 93–100

    2023

  21. [29]

    Privacy-preserving traffic flow prediction: A federated learning approach,

    Y . Liu et al. , “Privacy-preserving traffic flow prediction: A federated learning approach,” IEEE Internet of Things Journal , vol. 7, no. 8, pp. 7751–7763, 2020

    2020

  22. [30]

    Multi-task federated learning for traffic prediction and its application to route planning,

    T. Zeng et al., “Multi-task federated learning for traffic prediction and its application to route planning,” in 2021 IEEE Intelligent Vehicles Symposium (IV). IEEE, 2021, pp. 451–457

    2021

  23. [31]

    Dual attention-based federated learning for wireless traffic prediction,

    C. Zhang et al. , “Dual attention-based federated learning for wireless traffic prediction,” in IEEE INFOCOM 2021-IEEE conference on com- puter communications. IEEE, 2021, pp. 1–10

    2021

  24. [32]

    Efficient wireless traffic prediction at the edge: A federated meta-learning approach,

    L. Zhang, C. Zhang, and B. Shihada, “Efficient wireless traffic prediction at the edge: A federated meta-learning approach,”IEEE Communications Letters, vol. 26, no. 7, pp. 1573–1577, 2022

    2022

  25. [33]

    A multi-source dataset of urban life in the city of Milan and the Province of Trentino,

    G. Barlacchi et al., “A multi-source dataset of urban life in the city of Milan and the Province of Trentino,” Scientific data , vol. 2, no. 1, pp. 1–15, 2015

    2015

  26. [34]

    Federated learning: Challenges, methods, and future directions,

    T. Li et al. , “Federated learning: Challenges, methods, and future directions,” IEEE signal processing magazine, vol. 37, no. 3, pp. 50–60, 2020

    2020

  27. [35]

    Maverick matters: Client contribution and selection in federated learning,

    J. Huang et al., “Maverick matters: Client contribution and selection in federated learning,” in Pacific-Asia Conference on Knowledge Discovery and Data Mining . Springer, 2023, pp. 269–282

    2023

  28. [36]

    Toward understanding the influence of individual clients in federated learning,

    Y . Xue et al., “Toward understanding the influence of individual clients in federated learning,” in Proceedings of the AAAI Conference on Artificial Intelligence, vol. 35, no. 12, 2021, pp. 10 560–10 567

    2021

  29. [37]

    A deep-learning model for urban traffic flow prediction with traffic events mined from twitter,

    A. Essien et al., “A deep-learning model for urban traffic flow prediction with traffic events mined from twitter,” World Wide Web, vol. 24, no. 4, pp. 1345–1368, 2021

    2021

  30. [38]

    Graph attention spatial-temporal network with collabora- tive global-local learning for citywide mobile traffic prediction,

    K. He et al., “Graph attention spatial-temporal network with collabora- tive global-local learning for citywide mobile traffic prediction,” IEEE Transactions on mobile computing, vol. 21, no. 4, pp. 1244–1256, 2020

    2020

  31. [39]

    The Cost of Training Machine Learning Models over Distributed Data Sources,

    E. Guerra et al., “The Cost of Training Machine Learning Models over Distributed Data Sources,” IEEE Open Journal of the Communications Society, 2023

    2023

  32. [40]

    Long short-term memory,

    S. Hochreiter and J. Schmidhuber, “Long short-term memory,” Neural Computation, vol. 9, no. 8, 1997

    1997

  33. [41]

    3GPP, Tech. Specif. Group Radio Access Network; Physical layer procedures (Release 9), 3GPP TS 36.213

  34. [42]

    Carbontracker: Tracking and predicting the carbon footprint of training deep learning models,

    L. F. W. Anthony, B. Kanding, and R. Selvan, “Carbontracker: Tracking and predicting the carbon footprint of training deep learning models,” arXiv preprint arXiv:2007.03051 , 2020

    2007 arXiv

  35. [43]

    Isolation forest,

    F. T. Liu, K. M. Ting, and Z.-H. Zhou, “Isolation forest,” in 2008 eighth ieee international conference on data mining. IEEE, 2008, pp. 413–422

    2008

  36. [44]

    Alternatives to the median absolute deviation,

    P. J. Rousseeuw and C. Croux, “Alternatives to the median absolute deviation,” Journal of the American Statistical association , vol. 88, no. 424, pp. 1273–1283, 1993

    1993

  37. [45]

    Robust outlier detection using svm regression,

    E. M. Jordaan and G. F. Smits, “Robust outlier detection using svm regression,” in IEEE International Joint Conference on Neural Networks, vol. 3, 2004, pp. 2017–2022

    2004

  38. [46]

    A review on outlier/anomaly detection in time series data,

    A. Bl ´azquez-Garc´ıa et al. , “A review on outlier/anomaly detection in time series data,” ACM computing surveys (CSUR) , vol. 54, no. 3, pp. 1–33, 2021

    2021

  39. [47]

    Tackling the objective inconsistency problem in het- erogeneous federated optimization,

    J. Wang et al. , “Tackling the objective inconsistency problem in het- erogeneous federated optimization,” Advances in neural information processing systems, vol. 33, pp. 7611–7623, 2020

    2020

  40. [48]

    Federated visual classification with real-world data distribution,

    T.-M. H. Hsu, H. Qi, and M. Brown, “Federated visual classification with real-world data distribution,” in Computer Vision–ECCV 2020: 16th European Conference, Glasgow, UK, August 23–28, 2020, Proceedings, Part X 16. Springer, 2020, pp. 76–92

    2020

  41. [49]

    Adaptive federated optimization,

    S. Reddi et al. , “Adaptive federated optimization,” arXiv preprint arXiv:2003.00295, 2020

    2003 arXiv

  42. [50]

    Amplitude-aligned personalization and robust aggregation for federated learning,

    Y . Jiang, S. Chen, and X. Bao, “Amplitude-aligned personalization and robust aggregation for federated learning,” IEEE Transactions on Sustainable Computing, 2023

    2023

  43. [51]

    Upgini - automated data search & enrichment library,

    R. Post and N. Toro, “Upgini - automated data search & enrichment library,” Jul. 2022. [Online]. Available: https://github.com/upgini/upgini 12 Nikolaos Pavlidis is a PhD student at the Dept. of Electrical and Computer Engineering of the Dem- ocritus University of Thrace (Gree...

    2020

Pith tools

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