Pith. sign in

REVIEW 5 major objections 6 minor 37 references

FedPAW: Federated Learning with Personalized Aggregation Weights for Urban Vehicle Speed Prediction

T0 review · 5 major / 6 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 FedPAW claims that server-side personalized aggregation weights can tailor federated speed-prediction models to individual drivers without adding any client-side computation or communication, and reports the lowest 10-second-horizon error…

desk verdict FedPAW is a sensible server-side personalization heuristic with a useful new dataset and code, but the headline 0.8% improvement over FedRep is within noise and the hyperparameters are tuned on the test set. read the letter →

arxiv 2412.01281 v1 pith:SIXTU3UE submitted 2024-12-02 cs.AI cs.DC

Yuepeng He , Pengzhan Zhou , Yijun Zhai , Fang Qu , Zhida Qin , Mingyan Li , Songtao Guo This is my paper · ORCID
classification cs.AIcs.DC
keywords federatedlearningpersonalizedvehiclespeedpredictionaggregationweightsCARLAsimulatorSeq2SeqLSTMmulti-headattentionnon-IIDdata
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

FedPAW sets out to show that personalization in federated learning does not have to cost the client anything: a server that already aggregates local models can create a distinct model for each vehicle by weighting the blend of global and local parameters according to how much those parameters differ across clients. The paper's central claim is that the parameters that diverge most after federated averaging encode driver- and vehicle-specific behavior, so amplifying them in the returned model yields personalized speed predictions. On a ten-client simulated urban driving dataset, the method ranks lowest in prediction error at the 10-second horizon, cutting test MAE by 0.8% relative to the best of eleven baselines, with the same client-side computation and communication as FedAvg. This matters for intelligent transportation because drivers' styles and vehicle types make speed prediction inherently personal, while vehicles have limited resources for extra local fine-tuning.

What carries the argument

The personalized aggregation weight tensor $W^{t,p}$ is the central object. It is computed once per round on the server from the weighted mean squared error between the top-$p$ layers of the participating clients' local models and the global model, then min-max normalized layer by layer so every weight lies in $[0,1]$. Multiplying the local-global parameter difference by these weights decides, element by element, how much personalized information from the client's own model survives in the model the server sends back. A hyperparameter $r$ keeps the weights at zero for the first rounds, so FedPAW starts as FedAvg, and a hyperparameter $p$ restricts personalization to the upper layers on the grounds that lower layers encode general knowledge.

What would settle it

Shuffle the driver labels in CarlaVSP so each client's local data no longer corresponds to a consistent driver, retrain FedPAW and FedAvg from the same initialization, and compare test MAE; if FedPAW still beats FedAvg, the reported 0.8% gain cannot be attributed to the personalized weights. A second check is to permute the entries of $W^{t,p}$ randomly at each round: if test error is unchanged, the specific weighting is not what matters.

Watch

Extended reading notes

Core claim

FedPAW claims that a server, after averaging client models as in FedAvg, can build each client a personalized model by element-wise blending the global model with that client's local model. The blend is $\hat{\Theta}_i^{t+1} = \Theta^t + (\Theta_i^t - \Theta^t) \odot [0_{L-p}; W^{t,p}]$, so only the top $p$ layers are personalized while lower layers keep the shared general features. The weight tensor $W^{t,p}$ is the min-max normalized, data-size-weighted mean squared difference between the top-$p$ parameters of the selected local models and the global model. The paper argues that parameters that diverge most across clients are exactly the parameters carrying personalized driving information, and that amplifying them on the server yields models tailored to each driver and vehicle. On the ten-client CarlaVSP dataset, FedPAW reports the lowest test MAE and RMSE at the 10-second prediction horizon among eleven baselines, with a 0.8% lower MAE and a 0.9% lower RMSE than the next-best method at full client participation; at 5 seconds FedRep is slightly better but requires extra client-side fine-tuning.

Load-bearing premise

The load-bearing premise is that after federated averaging, model parameters that differ most between clients carry exactly the personalized information the method should amplify; if divergence is driven instead by noise, initialization, or training instability, the aggregation weights will not improve prediction.

Editorial extensions

If this is right

  • If FedPAW is right, vehicle speed prediction can be personalized under federated privacy constraints without any added client-side computation or communication, since all personalization is a few matrix operations on the server.
  • The personalization mechanism is independent of the specific predictor: the same top-$p$ weighted blending can be applied to any neural speed predictor trained under FedAvg-style federated learning.
  • At the 10-second horizon the reported advantage persists under dynamic client participation ($\rho \in [0.1,1]$), so the method is stable when vehicles drop in and out of the training process.
  • Because FedAvg is the special case with all aggregation weights zero, FedPAW can only depart from FedAvg where the data say the parameters differ, giving it a graceful fallback.

Reading between the lines

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

  • A testable extension is to apply FedPAW to real-world multi-driver trajectory logs; if real driver heterogeneity is stronger than the simulator's, the 0.8% gain could be larger, and if weaker, it could vanish.
  • The min-max normalization is recomputed every round from the current client subset, so averaging or smoothing $W^{t,p}$ across rounds could reduce variance under partial participation, something the paper does not explore.
  • Comparing FedPAW to FedAvg on synthetic federated benchmarks with tunable non-IID degree would show whether the reported gain is a floor or a ceiling for the method's benefit.
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 proposes FedPAW, a personalized federated learning method for urban vehicle speed prediction. The server computes per-client aggregation weights from the weighted mean squared error between local model parameters and the global model, normalizes them layer-wise, and mixes the global model with each client's local model over the top p layers (Eqs. 4-7). The method adds no client-side computation or communication overhead beyond FedAvg. The authors collect a CARLA-based driving dataset (CarlaVSP) with 10 clients, propose an LSTM Seq2Seq model with multi-head attention, and report that FedPAW achieves the lowest MAE at the 10-second prediction horizon, with a 0.8% reduction in test MAE over the best baseline (FedRep), while being slightly worse than FedRep at 5 seconds. The paper also includes feature-group analysis, hyperparameter sensitivity, overhead measurements, and qualitative prediction plots.

Significance. If the central empirical claim holds, the paper makes a useful contribution: it shows that server-side personalized aggregation can improve prediction accuracy without burdening clients, and it releases an open dataset and code, which are valuable assets for future work on personalized vehicle speed prediction. The paper is also careful to compare against eleven baselines and to report computation/communication overhead. However, the significance of the headline result is currently limited by the absence of statistical significance testing and by the apparent selection of feature groups and hyperparameters on the test setup. The method itself is heuristic, and the paper does not provide an ablation that isolates the contribution of the proposed weighting scheme, so the claimed advantage is not yet convincingly established beyond what could be attributed to random variation or selection effects.

major comments (5)
  1. [Table VI, Section V-D] The central claim rests on Table VI, 10-s horizon, rho=1: FedPAW MAE 1.635 +/- 0.011 vs. FedRep 1.648 +/- 0.008, both with five runs. No significance test (paired or unpaired) is reported, and the standard deviations overlap; at the 5-s horizon the ranking reverses (FedRep 1.147 vs. FedPAW 1.163). The statement that FedPAW 'ranks lowest' is therefore not statistically supported as written. Please provide per-run or per-client paired comparisons, confidence intervals, or a significance test; if the 10-s gap is not significant, the conclusion should be softened accordingly.
  2. [Section V-C, Table IV] The hyperparameter p=4 for the 10-s horizon is said to be chosen 'based on similar experiments,' but no 10-s sweep over p is shown; Table IV reports only 5-s results, and r=1 is also selected from 5-s experiments and then applied to 10-s. Since the claimed advantage is at 10-s, this constitutes potential selection on the test setup unless a separate validation split was used. Please report the 10-s hyperparameter sweep or describe an explicit nested validation procedure that prevents test-set information from influencing p and r.
  3. [Section V-B, Table III] Feature group FG6 is selected as the best-performing group on the basis of the same test-evaluation protocol (Table III) and is then used for all subsequent benchmark comparisons. This is a form of test-set selection that can inflate the reported improvements of the whole pipeline. Please either select features on a validation split (e.g., a held-out portion of each client's training data) or report the key FedPAW-vs-baseline comparison for all feature groups to show that the conclusion is robust to feature choice.
  4. [Section V-A, Table VI] The benchmark set is dominated by client-side personalization methods (Per-FedAvg, FedRep, pFedMe, Ditto, APFL, FedFomo, FedALA). Since FedPAW's stated novelty is server-side personalization without client overhead, the comparison should include server-side personalization baselines such as FedBN, LG-FedAvg, or a server-side variant of FedALA that does not require client-side adaptation layers. Without such baselines, the claim that FedPAW is the best server-side personalized aggregation method is not established.
  5. [Section IV-C, Eqs. (6)-(7)] The aggregation-weight design is justified only by an 'empirical assumption' that parameter differences indicate personalized information. The paper does not demonstrate that the specific weighted-MSE plus min-max normalization is necessary or superior to simpler alternatives (e.g., uniform mixing over the top p layers, or a scalar mixing coefficient per client). An ablation that varies the weight computation would substantiate the causal claim that the proposed weighting mechanism, rather than the mere act of mixing local and global models, drives the reported improvement.
minor comments (6)
  1. [Section III-A] Typos: 'selectino' should be 'selection' and 'approporiate' should be 'appropriate' in the opening paragraph.
  2. [Section IV-C, Eq. (4)] The notation W^t in Eq. (4) is used without the layer-range superscript p, while Eq. (5) introduces W^{t,p}; please make the notation consistent throughout the derivation.
  3. [Algorithm 1] Line 14 returns models only for clients in S_t; it is unclear what model is used for clients not sampled in a given iteration. Please clarify how the final returned models are obtained for all clients.
  4. [Figure 8] The convergence curves in Figure 8 are shown without error bars or confidence bands; adding them would help the reader assess the variability of the MAE trajectories.
  5. [Section V-E] The term 'stability' is used to describe robustness to varying client joining ratios; consider using a more specific term such as 'participation robustness' to avoid confusion with statistical stability.
  6. [Abstract and Section I] The '0.8% reduction' is computed from the difference between 1.635 and 1.648; stating the absolute values in the abstract would help readers judge the practical significance of the gain.

Circularity Check

0 steps flagged · score 0.0 of 10

No circular derivation: FedPAW's aggregation weights are a heuristic computed from the same local models they mix, but the claimed 0.8% MAE advantage is an empirical benchmark result, not a quantity forced by construction.

full rationale

FedPAW's derivation chain is algorithmic rather than inferential. Equations (6) and (7) define the aggregation weights W^{t,p} as a min-max normalized weighted mean squared parameter difference between the client local models and the global model; Equation (5) then uses those weights to interpolate per-coordinate between the global model and each client's local model. This is a well-defined heuristic: the weights are computed from the same local models that are later mixed, but the paper does not claim that Equation (6) or (7) mathematically forces the reported test MAE. The 'empirical assumption' in Section IV-C (that large local-global parameter differences indicate personalized information) is explicitly labeled an assumption, and the paper's headline result is validated empirically against eleven external baselines rather than derived from that assumption. The 10-s hyperparameter p=4 is selected experimentally ('adjust p=4 for the prediction horizon of 10 s based on similar experiments'), which is standard model selection rather than a fitted parameter renamed as a prediction; no equation reduces the 0.8% margin to a fitted constant. The only apparent self-citation is reference [24] (Z. Zhou among the authors), used to support the general statement that a single global model struggles under heterogeneity; this is not load-bearing for the FedPAW derivation. No uniqueness theorem, ansatz-via-citation, or renaming of a known result is used. The central claim may have statistical-support limitations (no significance test, overlapping standard deviations at the 10-s horizon), but those are correctness risks, not circularity.

Assumptions & free parameters 2 free parameters · 2 assumptions · 0 invented entities

The method's central claim rests on two empirical assumptions about model parameters and layer roles, plus two tuned hyperparameters (p and r). No new physical entities are introduced. The free parameters are chosen from the same experimental setup used to report the headline result.

free parameters (2)
  • p (number of top layers for personalized aggregation) = 2 for 5s, 4 for 10s
    Controls how many top layers are mixed with local parameters; chosen in Section V-C by scanning p=1..max and selecting the lowest test error.
  • r (iterations before personalization starts) = 1
    Rounds of pure FedAvg before PA; the authors report r=1 is best because the task is simple and models converge quickly.
assumptions (2)
  • domain assumption After several FedAvg rounds, parameters that differ strongly across client local models encode personalized information, so per-layer variance can be used as aggregation weights.
    Stated in Section IV-C as an empirical assumption; used to justify Eqs. (6)-(7).
  • domain assumption The top p layers of a network carry more personalized information and lower layers carry general information, so mixing only the top p layers is sufficient.
    Invoked in Section IV-C, based on reference [33]; underlies the p hyperparameter.

how reviews work

0 comments
Cite this review

Pith. "Pith review of FedPAW: Federated Learning with Personalized Aggregation Weights for Urban Vehicle Speed Prediction." pith.science (2026). https://pith.science/paper/SIXTU3UE

@misc{pith2026241201281,
  author       = {Pith},
  title        = {Pith review of: FedPAW: Federated Learning with Personalized Aggregation Weights for Urban Vehicle Speed Prediction},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/SIXTU3UE}},
  note         = {Machine review of arXiv:2412.01281}
}
read the original abstract

Vehicle speed prediction is crucial for intelligent transportation systems, promoting more reliable autonomous driving by accurately predicting future vehicle conditions. Due to variations in drivers' driving styles and vehicle types, speed predictions for different target vehicles may significantly differ. Existing methods may not realize personalized vehicle speed prediction while protecting drivers' data privacy. We propose a Federated learning framework with Personalized Aggregation Weights (FedPAW) to overcome these challenges. This method captures client-specific information by measuring the weighted mean squared error between the parameters of local models and global models. The server sends tailored aggregated models to clients instead of a single global model, without incurring additional computational and communication overhead for clients. To evaluate the effectiveness of FedPAW, we collected driving data in urban scenarios using the autonomous driving simulator CARLA, employing an LSTM-based Seq2Seq model with a multi-head attention mechanism to predict the future speed of target vehicles. The results demonstrate that our proposed FedPAW ranks lowest in prediction error within the time horizon of 10 seconds, with a 0.8% reduction in test MAE, compared to eleven representative benchmark baselines. The source code of FedPAW and dataset CarlaVSP are open-accessed at: https://github.com/heyuepeng/PFLlibVSP and https://pan.baidu.com/s/1qs8fxUvSPERV3C9i6pfUIw?pwd=tl3e.

Figures

Figures reproduced from arXiv: 2412.01281 by the authors.

Figure 1
Figure 1. Illustration of FedPAW on the server. Federated Learning (FL) emerges as a collaborative, dis￾tributed machine learning paradigm coordinated by a central server and solved jointly through a network of multiple par￾ticipating devices (clients). Each client possesses its private dataset, not uploaded to the server. Instead, locally trained models are uploaded to aggregate into a global model on the server, effectively… view at source ↗
Figure 2
Figure 2. Data collection scenarios. (a): Urban map Town 10 [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 4
Figure 4. Schematic of V2V and V2I enabled traffic scenario. [PITH_FULL_IMAGE:figures/full_fig_p004_4.png] view at source ↗
Figures from the paper (7 more)
Figure 3
Figure 3. Figure 3: Vehicle speed distribution of 10 clients during driving [PITH_FULL_IMAGE:figures/full_fig_p004_3.png]
Figure 5
Figure 5. Figure 5: Multi-Head Attention Augmented Seq2Seq LSTM [PITH_FULL_IMAGE:figures/full_fig_p005_5.png]
Figure 6
Figure 6. Figure 6: Overview of the FedPAW Framework: ① Clients upload local models to the server. ② The server performs personalized aggregation using the client local models and the global model to produce the aggregated models. ③ The server sends tailored aggregated models to the clien…
Figure 7
Figure 7. Figure 7: illustrates the computation process for the top p layers of Personalized Aggregation Weights. Algorithm 1 describes the entire FL process within FedPAW. Θ𝑖 𝑡,𝑝 … … The top 𝒑 layers of local models for clients and global model … … … … … … 𝑊𝑡,𝑝 Θ 𝑡,𝑝 𝑀𝑡,𝑝 Θ𝑖 𝑡,𝑝 Θ 𝑡,𝑝 No…
Figure 8
Figure 8. Figure 8: Test MAE (m/s) curves of different methods at a [PITH_FULL_IMAGE:figures/full_fig_p009_8.png]
Figure 9
Figure 9. Figure 9: Speed prediction results of FedPAW framework on a subset of the test dataset at a prediction horizon of 5 s. [PITH_FULL_IMAGE:figures/full_fig_p010_9.png]
Figure 10
Figure 10. Figure 10: Speed prediction results results of the CV model, CA [PITH_FULL_IMAGE:figures/full_fig_p010_10.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

37 extracted references · 18 canonical work pages

  1. [1]

    Vehicle speed prediction by two-level data driven models in vehicular networks,

    B. Jiang and Y . Fei, “Vehicle speed prediction by two-level data driven models in vehicular networks,” IEEE Transactions on Intelligent Transportation Systems, vol. 18, no. 7, pp. 1793–1801, 2016. IEEE TRANSACTIONS ON CLOUD COMPUTING, XX 11

    work page 2016

  2. [2]

    Velocity predictors for predic- tive energy management in hybrid electric vehicles,

    C. Sun, X. Hu, S. J. Moura, and F. Sun, “Velocity predictors for predic- tive energy management in hybrid electric vehicles,” IEEE Transactions on Control Systems Technology , vol. 23, no. 3, pp. 1197–1204, 2014

    work page 2014

  3. [3]

    Short-term vehicle speed prediction based on convolutional bidirectional lstm networks,

    S. Han, F. Zhang, J. Xi, Y . Ren, and S. Xu, “Short-term vehicle speed prediction based on convolutional bidirectional lstm networks,” in 2019 IEEE intelligent transportation systems conference (ITSC). IEEE, 2019, pp. 4055–4060

    work page 2019

  4. [4]

    Short term prediction of a vehicle’s velocity trajectory using its,

    D. Moser, H. Waschl, R. Schmied, H. Efendic, and L. Del Re, “Short term prediction of a vehicle’s velocity trajectory using its,” SAE Inter- national Journal of Passenger Cars-Electronic and Electrical Systems , vol. 8, no. 2015-01-0295, pp. 364–370, 2015

    work page 2015

  5. [5]

    An lstm-based speed predictor based on traffic simulation data for improving the performance of energy-optimal adaptive cruise control,

    Y . Jia, C. Cai, and D. G ¨orges, “An lstm-based speed predictor based on traffic simulation data for improving the performance of energy-optimal adaptive cruise control,” in 2020 IEEE 23rd International Conference on Intelligent Transportation Systems (ITSC) . IEEE, 2020, pp. 1–7

    work page 2020

  6. [6]

    Deep learning- based vehicle speed prediction for ecological adaptive cruise control in urban and highway scenarios,

    S. K. Chada, D. G ¨orges, A. Ebert, and R. Teutsch, “Deep learning- based vehicle speed prediction for ecological adaptive cruise control in urban and highway scenarios,” IFAC-PapersOnLine, vol. 56, no. 2, pp. 1107–1114, 2023

    work page 2023

  7. [7]

    Stochastic model predictive control with driver behavior learning for improved powertrain control,

    M. Bichi, G. Ripaccioli, S. Di Cairano, D. Bernardini, A. Bemporad, and I. V . Kolmanovsky, “Stochastic model predictive control with driver behavior learning for improved powertrain control,” in 49th IEEE conference on decision and control (CDC). IEEE, 2010, pp. 6077–6082

    work page 2010

  8. [8]

    Predictive cruise control with probabilistic constraints for eco driving,

    C. Zhang and A. Vahidi, “Predictive cruise control with probabilistic constraints for eco driving,” in Dynamic Systems and Control Confer- ence, vol. 54761, 2011, pp. 233–238

    work page 2011

Show all 37 references
  1. [9]

    Prediction of human driving behavior using dynamic bayesian networks,

    T. Kumagai and M. Akamatsu, “Prediction of human driving behavior using dynamic bayesian networks,” IEICE TRANSACTIONS on Infor- mation and Systems , vol. 89, no. 2, pp. 857–860, 2006

    2006

  2. [10]

    Modeling and recognizing driver behavior based on driving data: A survey,

    W. Wang, J. Xi, H. Chen et al. , “Modeling and recognizing driver behavior based on driving data: A survey,” Mathematical Problems in Engineering, vol. 2014, 2014

    2014

  3. [11]

    Longitudinal vehicle motion prediction in urban settings with traffic light interaction,

    M. Wegener, F. Herrmann, L. Koch, R. Savelsberg, and J. Andert, “Longitudinal vehicle motion prediction in urban settings with traffic light interaction,” IEEE Transactions on Intelligent Vehicles , vol. 8, no. 1, pp. 204–215, 2021

    2021

  4. [12]

    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

  5. [13]

    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

  6. [14]

    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, 2022

    2022

  7. [15]

    Fedala: Adaptive local aggregation for personalized federated learning,

    J. Zhang, Y . Hua, H. Wang, T. Song, Z. Xue, R. Ma, and H. Guan, “Fedala: Adaptive local aggregation for personalized federated learning,” in Proceedings of the AAAI Conference on Artificial Intelligence, vol. 37, no. 9, 2023, pp. 11 237–11 244

    2023

  8. [16]

    Carla: An open urban driving simulator,

    A. Dosovitskiy, G. Ros, F. Codevilla, A. Lopez, and V . Koltun, “Carla: An open urban driving simulator,” in Conference on robot learning . PMLR, 2017, pp. 1–16

    2017

  9. [17]

    Personalized federated learning with theoretical guarantees: A model-agnostic meta-learning approach,

    A. Fallah, A. Mokhtari, and A. Ozdaglar, “Personalized federated learning with theoretical guarantees: A model-agnostic meta-learning approach,” Advances in Neural Information Processing Systems, vol. 33, pp. 3557–3568, 2020

    2020

  10. [18]

    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

  11. [19]

    Personalized federated learning with moreau envelopes,

    C. T Dinh, N. Tran, and J. Nguyen, “Personalized federated learning with moreau envelopes,” Advances in Neural Information Processing Systems, vol. 33, pp. 21 394–21 405, 2020

    2020

  12. [20]

    Ditto: Fair and robust federated learning through personalization,

    T. Li, S. Hu, A. Beirami, and V . Smith, “Ditto: Fair and robust federated learning through personalization,” in International conference on machine learning . PMLR, 2021, pp. 6357–6368

    2021

  13. [21]

    Adaptive personalized federated learning,

    Y . Deng, M. M. Kamani, and M. Mahdavi, “Adaptive personalized federated learning,” arXiv preprint arXiv:2003.13461 , 2020

    2003 arXiv

  14. [22]

    Person- alized federated learning with first order model optimization,

    M. Zhang, K. Sapra, S. Fidler, S. Yeung, and J. M. Alvarez, “Person- alized federated learning with first order model optimization,” arXiv preprint arXiv:2012.08565, 2020

    2012 arXiv

  15. [23]

    Advances and open problems in federated learning,

    P. Kairouz, H. B. McMahan, B. Avent, A. Bellet, M. Bennis, A. N. Bhagoji, K. Bonawitz, Z. Charles, G. Cormode, R. Cummings et al. , “Advances and open problems in federated learning,” Foundations and trends® in machine learning , vol. 14, no. 1–2, pp. 1–210, 2021

    2021

  16. [24]

    Per- sonalized cross-silo federated learning on non-iid data,

    Y . Huang, L. Chu, Z. Zhou, L. Wang, J. Liu, J. Pei, and Y . Zhang, “Per- sonalized cross-silo federated learning on non-iid data,” in Proceedings of the AAAI conference on artificial intelligence , vol. 35, no. 9, 2021, pp. 7865–7873

    2021

  17. [25]

    Federated meta-learning with fast convergence and efficient communication,

    F. Chen, M. Luo, Z. Dong, Z. Li, and X. He, “Federated meta-learning with fast convergence and efficient communication,” arXiv preprint arXiv:1802.07876, 2018

    2018 arXiv

  18. [26]

    Federated learning with personalization layers,

    M. G. Arivazhagan, V . Aggarwal, A. K. Singh, and S. Choud- hary, “Federated learning with personalization layers,” arXiv preprint arXiv:1912.00818, 2019

    1912 arXiv

  19. [27]

    Traffic light assistant system for optimized energy consumption in an electric vehicle,

    E. Kural, S. Jones, A. F. Parrilla, and A. Grauers, “Traffic light assistant system for optimized energy consumption in an electric vehicle,” in2014 International Conference on Connected Vehicles and Expo (ICCVE) . IEEE, 2014, pp. 604–611

    2014

  20. [28]

    Safe- and eco-driving control for connected and automated electric vehicles using analytical state-constrained optimal solution,

    J. Han, A. Sciarretta, L. L. Ojeda, G. De Nunzio, and L. Thibault, “Safe- and eco-driving control for connected and automated electric vehicles using analytical state-constrained optimal solution,” IEEE Transactions on Intelligent Vehicles, vol. 3, no. 2, pp. 163–172, 2018

    2018

  21. [29]

    Vehicle speed prediction with rnn and attention model under multiple scenarios,

    C.-S. Shih, P.-W. Huang, E.-T. Yen, and P.-K. Tsung, “Vehicle speed prediction with rnn and attention model under multiple scenarios,” in 2019 IEEE Intelligent Transportation Systems Conference (ITSC) . IEEE, 2019, pp. 369–375

    2019

  22. [30]

    Short-term vehicle speed prediction based on bilstm-gru model considering driver heterogeneity,

    Q. Li, R. Cheng, and H. Ge, “Short-term vehicle speed prediction based on bilstm-gru model considering driver heterogeneity,” Physica A: Statistical Mechanics and its Applications , vol. 610, p. 128410, 2023

    2023

  23. [31]

    Learning phrase representations using rnn encoder-decoder for statistical machine translation,

    K. Cho, B. Van Merri ¨enboer, C. Gulcehre, D. Bahdanau, F. Bougares, H. Schwenk, and Y . Bengio, “Learning phrase representations using rnn encoder-decoder for statistical machine translation,” arXiv preprint arXiv:1406.1078, 2014

    2014 arXiv

  24. [32]

    Attention is all you need,

    A. Vaswani, N. Shazeer, N. Parmar, J. Uszkoreit, L. Jones, A. N. Gomez, Ł. Kaiser, and I. Polosukhin, “Attention is all you need,” Advances in neural information processing systems , vol. 30, 2017

    2017

  25. [33]

    Data-free knowledge distillation for het- erogeneous federated learning,

    Z. Zhu, J. Hong, and J. Zhou, “Data-free knowledge distillation for het- erogeneous federated learning,” in International conference on machine learning. PMLR, 2021, pp. 12 878–12 889

    2021

  26. [34]

    Adam: A method for stochastic optimization,

    D. P. Kingma and J. Ba, “Adam: A method for stochastic optimization,” arXiv preprint arXiv:1412.6980 , 2014. Yuepeng He received the bachelor’s degree in software engineering from Nanchang University, in

    2014 arXiv

  27. [36]

    His research in- terests in federated learning and internet of vehicles

    He is currently working toward the master’s degree with Chongqing University. His research in- terests in federated learning and internet of vehicles. Fang Qu received the bachelor’s degree in computer science from Chongqing University,in 2023. He is currently working toward t...

    2023

  28. [2020]

    She joined the College of Computer Science, Chongqing University, in 2021, and currently works as a Post-Doctoral Research Associate. Her current research interests include industrial wireless networks and application in industrial automation, joint design of communication and...

    2021

  29. [2022]

    His research in- terests in personalized federated learning and inter- net of vehicles

    He is currently working toward the master’s degree with Chongqing University. His research in- terests in personalized federated learning and inter- net of vehicles. Pengzhan Zhou received the BS degree in both ap- plied physics and applied mathematics from Shang- hai Jiaotong...

Pith tools

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