Pith. sign in

REVIEW 3 major objections 5 minor 49 references

Optimized Local Updates in Federated Learning via Reinforcement Learning

T0 review · 3 major / 5 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read A deep reinforcement learning agent that chooses per-class data proportions during federated rounds, followed by fine-tuning on the full local dataset, can make a client outperform naive federated clients on multiple benchmarks and…

desk verdict A reasonable RL-for-FL idea is undermined in its current form by a confounded main comparison and a tautological theorem; the claims need a fine-tuning control before they can be taken seriously. read the letter →

arxiv 2506.06337 v1 pith:MGQGSU5H submitted 2025-05-31 cs.LG

Ali Murad , Bo Hui , Wei-Shinn Ku This is my paper
classification cs.LG
keywords federatedlearningdeepreinforcementdataheterogeneitynon-IIDclientpersonalizationselectionlocalupdatesDDPG
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

Federated learning trains a shared model across clients without centralizing raw data, but non-IID local datasets can make aggregation hurt individual clients. This paper argues that, for a given client, training locally on more data than necessary does not help and can even hurt, so the client should let a deep reinforcement learning agent pick, each round, how much of each class's local data to train on. After the federated rounds, the client fine-tunes on its complete local dataset, and the paper reports that this optimized client outperforms naive clients on CIFAR-10, CIFAR-100, and FashionMNIST under five aggregation algorithms, for example 43.01% versus 26.64% accuracy on CIFAR-10 with FedAvg. The paper also proves an upper bound on the optimized client's performance gap relative to full-data training during the federated rounds, framing the approach as a trade-off between during-FL and post-FL performance. If correct, the method gives a client a way to personalize the global model while feeding less data into the server's aggregation.

What carries the argument

The load-bearing object is the DRL action vector $a_t = (z_1, \dots, z_C)$, where $z_c$ is the proportion of class-$c$ samples selected from the client's local dataset for the next FL round. The state is the class-wise F1-score of the server-aggregated model on the client's local validation split; the reward is a piecewise function of relative training-loss reduction, the mean action, and a fitted loss-estimation curve; and the policy is trained with DDPG. Two exploration variants, the normalized action and the weighted-metric action, adapt the action vector to epsilon-greedy selection. The accompanying bound, $P_k - P'_k \le \Omega$, is a circle-area argument: performance for a class is treated as the area of a circle whose radius is the fraction of data used, so the gap between full-data training and partitioned training is bounded by the sum of squared-radius differences. This bound is what lets the client tolerate lower during-FL accuracy while expecting post-FL gains.

What would settle it

Run the same federated setup with three clients: an optimized client, a naive client that trains on all local data throughout, and a control client that trains on all local data during FL and then performs the same post-FL full-data fine-tuning without any DRL agent. If the control client's final accuracy matches or beats the optimized client's, then the DRL data selection is not the source of the reported gain.

Watch

Extended reading notes

Core claim

The central claim is that a federated learning client can achieve better performance on its own data distribution by not using all its local data during the federated rounds. The proposed optimized client uses a DRL agent that observes the class-wise F1-score of the aggregated model on the client's local validation data as its state, outputs per-class data-selection weights as its action, and receives a reward based on the change in training loss and the amount of data used. The selected subset is used only for local FL training; after the final aggregation round, the client fine-tunes on its complete local dataset. Reported results compare the best mean performance of naive clients with the optimized client across FedAvg, FedAvgM, FedMedian, FedProx, and FedCDA, with the optimized client reaching 43.01% accuracy on CIFAR-10 versus 26.64% for naive FedAvg clients. The paper frames this as both a personalization mechanism and a way to avoid oversharing local information with the server.

Load-bearing premise

The load-bearing premise is that the measured performance gain of the optimized client comes from the DRL-chosen data partitions, but the optimized client alone also receives extra fine-tuning on its complete local dataset after the FL rounds, and no naive control client gets that same post-FL fine-tuning without the DRL agent.

Editorial extensions

If this is right

  • A client can improve its final personalized accuracy by letting a DRL agent decide per-class data proportions during federated rounds and then fine-tuning on its full local dataset after aggregation ends.
  • The reported improvement appears across all five tested aggregation algorithms, suggesting the data-selection mechanism does not depend on a particular aggregation rule.
  • During the federated rounds the optimized client may lag behind naive clients, but the upper-bound result says this lag is controlled by the gap between the selected action weights and the full per-class sample sizes.
  • Because the optimized client uses less local data in each round, it shares less information with the server during aggregation than a naive client, aligning personalization with privacy.
  • The ablation study indicates that replacing optimized actions with naive 80/20 actions changes the outcome, which the paper credits to the learned data-partitioning policy.

Reading between the lines

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

  • A control experiment that gives a naive client the same post-FL full-data fine-tuning without the DRL agent would isolate whether the reported gains come from the learned data selection or from the extra training phase; the current comparison does not run that control.
  • The per-class selection policy is defined for only one optimized client; if several clients ran DRL agents simultaneously, their choices would interact through the server aggregation, so the bound $P_k - P'_k \le \Omega$ would need a coupled multi-agent version rather than a per-client one.
  • Because the state is a class-wise performance vector and the reward is pluggable, the same machinery could be repurposed for other objectives such as class-balanced fairness or energy budgets, but the paper does not test those objectives.
  • A natural next test is whether the learned selection policy transfers across clients or datasets: train the agent on one client's data distribution and deploy it on another client in the same federation to see whether the selected proportions generalize.
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 proposes a DRL-based per-class data subsampling scheme for one "optimized" client in federated learning. The agent observes the client's F1-score after aggregation and outputs per-class weights that define how much of the local dataset is used in the next local-training round; the reward is a function of loss reduction and data usage. After the FL rounds, the optimized client fine-tunes on its full local dataset. Experiments compare the optimized client against naive clients on CIFAR-10, CIFAR-100, and FashionMNIST under five aggregation algorithms, and the paper presents a theoretical "performance upper bound" for the optimized client during FL.

Significance. If the claimed gains were attributable to the DRL data partition, the method would be a useful client-side personalization technique for non-IID federated learning. Strengths include the breadth of aggregation baselines, the availability of code, and a clearly described client-side algorithm. However, the main empirical comparison is confounded by an uncontrolled post-FL fine-tuning phase, and the theoretical analysis is a tautology. As presented, the results do not establish that the DRL policy provides any benefit, and the evidence in the ablation study points in the opposite direction during the FL rounds.

major comments (3)
  1. [Section V-C, Table I, Figure 4, Algorithm 1 lines 13-18] The headline comparison is confounded. Table I reports the optimized client's accuracy "after training on the complete local dataset, Dk, after the FL rounds," and Algorithm 1 lines 13-18 implement this extra fine-tuning phase. Naive clients do not receive an equivalent post-FL fine-tuning phase on their full local datasets. Therefore the reported gains (e.g., FedAvg accuracy 43.01% vs. 26.64%) conflate two interventions: the DRL-driven per-class data selection during FL and an additional local training stage on all local data. The ablation study in Table II makes the confound clear: during the FL rounds, the optimized client has lower accuracy than the baseline (FedAvg 26.91% vs. 34.98%). Without a control in which naive clients receive the same post-FL fine-tuning, or a report of the optimized client's accuracy immediately after round 100 without fine-tuning, the claim that the DRL partition itself yields superior performance is not established.
  2. [Section IV-B, Theorem 1] Theorem 1 is a tautology and provides no predictive content. Equations (10)-(12) define P_k and P'_k as sums of circle areas πZ_c^2 and πz_c^2, and Ω is defined as πΣ(Z_c^2 − z_c^2). The proof then derives P_k − P'_k = Ω by algebraic rearrangement. This is true by definition, not by any property of learning, loss, or generalization. The bound therefore says nothing about accuracy, loss, or any actual performance measure, and it does not support the paper's claim of an upper bound on client performance. The authors should either connect the circle-area quantity to a real performance metric through a nontrivial argument or remove the theorem and present the calculation as what it is: a definitional identity.
  3. [Section V-C, Figure 4, Table II, and the Abstract/Conclusion] The paper's own results contradict the unqualified claim of superior performance. Figure 4 and Table II show lower accuracy for the optimized client during the FL rounds under every aggregation algorithm, and the Discussion concedes that "training on a smaller partition of the local data can cause the optimized client to marginally lag in performance relative to the naive clients during the FL rounds." The abstract and conclusion nevertheless state that the method yields "superior performance" without the crucial qualification that this occurs only after an additional post-FL fine-tuning phase that is not applied to the baselines. This is a reporting issue that goes beyond style: it obscures the fact that the method's only demonstrated advantage comes from a stage that is absent from the comparison.
minor comments (5)
  1. [Section IV-A, Equation (5a)] Equation (5a) uses max(a'_c, |D_kc|)/|D_kc|, which cannot implement the stated goal of capping the selected count at the available class size; for any a'_c < |D_kc| the expression evaluates to 1, and for a'_c > |D_kc| it is also 1. The formula should use min(a'_c, |D_kc|)/|D_kc|.
  2. [Section IV, Equation (9a)] The condition in the reward function is written as "if T < τ," but T denotes the total number of FL rounds, not the current round index t; the intended condition is presumably "if t < τ."
  3. [Table I] In the FashionMNIST block, the FedCDA recall entry is printed as "57.91±4.59" without a percent sign, unlike all other entries; this is likely a typographical error.
  4. [Algorithm 1 and Section II-B] The parameter line "C ∈ (0, 1) 7→ R" is notationally nonsensical; it should be something like "C ∈ (0, 1)" for the client fraction. Also, "meausred" in Section II-B is a typo.
  5. [Figure 4 caption] The caption refers to a "dark green line" and "blue line," but it would be clearer to use the legend labels directly ("Optimized after FL," "Optimized during FL") given the multiple subplots.

Circularity Check

3 steps flagged · score 7.0 of 10

Theorem 1's 'performance upper bound' is an identity in the paper's own circle-area definitions; the reward includes the policy's own action; and Table I's claimed superiority is not isolated from an extra post-FL full-data fine-tuning phase.

  1. self definitional [Section IV-B, Theorem 1, Eqs. (10)-(13)]
    "The performance bound, of the client trained on the complete dataset, for class c is defined as the area of the circle: Pkc = πZ 2 c ... Pk − P ′ k = πZ 2 1 − πz 2 1 + πZ 2 2 − πz 2 2 + ·· + πZ 2 C − πz 2 C = πω1 + πω2 + πω3 + ·· + πωC ≤ π PC c=1 ωc = Ω"

    The 'performance' P_k is defined in Eq. (11) as π times the sum of squared class radii (sample-count ratios), and P'_k is defined similarly in Eq. (12). The proof then expands P_k − P'_k as π(Z_c^2 − z_c^2), calls each term πω_c, and defines Ω = πΣω_c. The theorem therefore restates the definitions: the upper bound is an exact equality by construction, not a derived bound on accuracy, loss, or generalization. It cannot predict client performance; it only re-expresses the chosen notational convention.

  2. self definitional [Section IV, Eq. (9a)-(9b) and surrounding text]
    "Rt ⇐ (Lagg − Ll)/Ll · 1/µat − λ if T < τ; (Lagg − Lest)/Lest · 1/µat − λ otherwise. Here ... µat = 1/|at| Σ at is the mean action generated by the policy πθ, and λ is a user defined parameter which normalizes the reward by controlling the amount of local training data generated by the policy."

    The reward used to train policy πθ includes μ_{a_t}, the mean of the action vector produced by that same policy. Because μ_{a_t} appears in the denominator, the agent can raise its reward by shrinking the magnitude of its own action even when the loss term is unchanged. The policy is therefore optimizing a target that contains its own output; part of the 'learning signal' is self-generated rather than an external measure of data-partition quality. The claimed optimization is self-referential by construction.

1 more flagged steps
  1. other [Section V-C, Table I, Table II, Algorithm 1 lines 13-18]
    "Each two-row combination shows a comparison of the mean performance achieved by all naive clients in the FL setup relative to the best performance of the optimized client, after training on the complete local dataset, Dk, after the FL rounds. ... It can be observed that the optimized client produces lower performance relative to the naive client during the FL rounds."

    Table I is the central evidence for 'superior performance.' Its optimized-column values are explicitly taken after the extra full-dataset training phase (Algorithm 1, lines 13-18), while the naive-client values are not. The same section reports that during FL rounds the optimized client is worse than naive clients. Hence the headline gain (e.g., FedAvg accuracy 43.01% vs. 26.64%) is not a prediction of the DRL partition; it is the trivial consequence of giving the optimized client a fine-tuning pass over Dk that the baseline never receives. Without a naive control that also fine-tunes on Dk after FL, the comparison reduces by construction to 'more training beats less training.'

full rationale

The paper is self-contained and does not rely on load-bearing self-citations; it benchmarks against standard FL baselines and includes an ablation study. However, the theoretical 'upper bound' (Theorem 1) is a tautology: 'performance' is defined as the circle area π times the squared sample-count radius, so P_k − P'_k is exactly Ω by definition, not a bound derived from learning dynamics. The reward function also feeds the policy's own mean action back into the reward, making the DRL objective partially self-referential: the agent can increase reward by shrinking its action magnitude independent of loss reduction. Finally, the empirical superiority claim is not isolated from the extra post-FL full-data fine-tuning: the paper itself concedes the optimized client is worse during FL rounds, and Table I reports only the post-fine-tuning numbers with no naive control receiving the same fine-tuning. Together these create partial circularity in the central claims, though not through any self-citation chain. Score 7 reflects that several central claims reduce by construction or by uncontrolled comparison, while the experimental benchmarks themselves remain externally grounded.

Assumptions & free parameters 3 free parameters · 2 assumptions · 0 invented entities

The method relies on a fitted loss curve for rewards, user-chosen hyperparameters, and two substantive axioms: that performance can be modeled as circle area from sample counts, and that more data always helps. The second axiom contradicts the paper's own motivating figure.

free parameters (3)
  • u, v (loss curve fit) = not reported
    Fitted per client to the local loss curve for the reward estimate L_est = -u * e^(-v * t); Section IV, Eq. 9b.
  • lambda = not reported
    User-defined reward normalization parameter controlling the data amount penalty; Section IV, Eq. 9a and 9b.
  • tau, eta, epsilon, b_l, b_u = not reported
    User-defined hyperparameters for loss estimation start, weighted-metric look-back, exploration, and action bounds; Section IV.
assumptions (2)
  • ad hoc to paper Client performance is represented as the area of a circle whose radius is the per-class sample count (P_kc = pi * Z_c^2).
    Introduced in Section IV-B Proposition 1 and used to derive Theorem 1; it is an invented equivalence with no link to actual model accuracy.
  • domain assumption Using more data leads to higher performance.
    Stated in Section IV-B as the basis for the bound; it conflicts with the paper's own motivation in Figure 1a that using all data can hurt.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Optimized Local Updates in Federated Learning via Reinforcement Learning." pith.science (2026). https://pith.science/paper/MGQGSU5H

@misc{pith2026250606337,
  author       = {Pith},
  title        = {Pith review of: Optimized Local Updates in Federated Learning via Reinforcement Learning},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/MGQGSU5H}},
  note         = {Machine review of arXiv:2506.06337}
}
read the original abstract

Federated Learning (FL) is a distributed framework for collaborative model training over large-scale distributed data, enabling higher performance while maintaining client data privacy. However, the nature of model aggregation at the centralized server can result in a performance drop in the presence of non-IID data across different clients. We remark that training a client locally on more data than necessary does not benefit the overall performance of all clients. In this paper, we devise a novel framework that leverages a Deep Reinforcement Learning (DRL) agent to select an optimized amount of data necessary to train a client model without oversharing information with the server. Starting without awareness of the client's performance, the DRL agent utilizes the change in training loss as a reward signal and learns to optimize the amount of training data necessary for improving the client's performance. Specifically, after each aggregation round, the DRL algorithm considers the local performance as the current state and outputs the optimized weights for each class, in the training data, to be used during the next round of local training. In doing so, the agent learns a policy that creates an optimized partition of the local training dataset during the FL rounds. After FL, the client utilizes the entire local training dataset to further enhance its performance on its own data distribution, mitigating the non-IID effects of aggregation. Through extensive experiments, we demonstrate that training FL clients through our algorithm results in superior performance on multiple benchmark datasets and FL frameworks. Our code is available at https://github.com/amuraddd/optimized_client_training.git.

Figures

Figures reproduced from arXiv: 2506.06337 by the authors.

Figure 1
Figure 1. Naive vs. Optimized clients. learns to optimize the amount of local training data used by the optimized client. The contributions of this paper are summarized as follows: • We provide a framework based on DRL to select local training data that a client uses to optimize its performance. Additionally, we investigate and present the results of our proposed framework using well-known FL aggregation algorithms. • We desi… view at source ↗
Figure 2
Figure 2. FL Workflow & RL Agent Environment Interaction. [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. Framework for optimized client training and action partitioned data selection. [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: Mean accuracy in FL rounds. The grey line shows the mean accuracy of all naive clients, whereas the blue line shows the accuracy of the optimized [PITH_FULL_IMAGE:figures/full_fig_p006_4.png]
Figure 5
Figure 5. Figure 5: Optimal actions taken by the RL agent in different federated learning [PITH_FULL_IMAGE:figures/full_fig_p007_5.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

49 extracted references · 32 canonical work pages

  1. [1]

    On the privacy-robustness-utility trilemma in distributed learning,

    Y . Allouah, R. Guerraoui, N. Gupta, R. Pinot, and J. Stephan, “On the privacy-robustness-utility trilemma in distributed learning,” in Proceed- ings of the 40th International Conference on Machine Learning , ser. Proceedings of Machine Learning Research, A. Krause, E. Brunskill, K. Cho, B. Engelhardt, S. Sabato, and J. Scarlett, Eds., vol. 202. PMLR, 23–...

    work page 2023

  2. [2]

    Flower: A friendly federated learning research framework,

    D. J. Beutel, T. Topal, A. Mathur, X. Qiu, J. Fernandez-Marques, Y . Gao, L. Sani, H. L. Kwing, T. Parcollet, P. P. d. Gusmão, and N. D. Lane, “Flower: A friendly federated learning research framework,”arXiv preprint arXiv:2007.14390, 2020

    arXiv 2007

  3. [3]

    API design for machine learning software: experiences from the scikit-learn project,

    L. Buitinck, G. Louppe, M. Blondel, F. Pedregosa, A. Mueller, O. Grisel, V . Niculae, P. Prettenhofer, A. Gramfort, J. Grobler, R. Layton, J. Van- derPlas, A. Joly, B. Holt, and G. Varoquaux, “API design for machine learning software: experiences from the scikit-learn project,” in ECML PKDD Workshop: Languages for Data Mining and Machine Learning , 2013, ...

    2013

  4. [4]

    Bayesian coreset optimization for personalized federated learning,

    P. Chanda, S. Modi, and G. Ramakrishnan, “Bayesian coreset optimization for personalized federated learning,” in The Twelfth International Conference on Learning Representations , 2024

    work page 2024

  5. [5]

    Calibre: Towards fair and accurate personalized federated learning with self-supervised learning,

    S. Chen, N. Su, and B. Li, “Calibre: Towards fair and accurate personalized federated learning with self-supervised learning,” in 2024 IEEE 44th International Conference on Distributed Computing Systems (ICDCS), 2024, pp. 891–901

    work page 2024

  6. [6]

    Momentum benefits non-iid federated learning simply and provably,

    Z. Cheng, X. Huang, P. Wu, and K. Yuan, “Momentum benefits non-iid federated learning simply and provably,” in The Twelfth International Conference on Learning Representations , 2024

    work page 2024

  7. [7]

    Muc-5 evaluation metrics,

    N. Chinchor and B. M. Sundheim, “Muc-5 evaluation metrics,” in Fifth Message Understanding Conference (MUC-5): Proceedings of a Conference Held in Baltimore, Maryland, August 25-27, 1993 , 1993

    work page 1993

  8. [8]

    Feddc: Federated learning with non-iid data via local drift decoupling and correction,

    L. Gao, H. Fu, L. Li, Y . Chen, M. Xu, and C.-Z. Xu, “Feddc: Federated learning with non-iid data via local drift decoupling and correction,” in Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2022, pp. 10 112–10 121

    work page 2022

Show all 49 references
  1. [9]

    Deep reinforcement learning for modelling protein complexes,

    Z. Gao, T. Feng, J. You, C. Zi, Y . Zhou, C. Zhang, and J. Li, “Deep reinforcement learning for modelling protein complexes,” in The Twelfth International Conference on Learning Representations , 2024

    2024

  2. [10]

    Goodfellow, Y

    I. Goodfellow, Y . Bengio, A. Courville, and Y . Bengio, Deep learning. MIT Press, 2016, vol. 1

    2016

  3. [11]

    Array programming with NumPy,

    C. R. Harris, K. J. Millman, S. J. van der Walt, R. Gommers, P. Virtanen, D. Cournapeau, E. Wieser, J. Taylor, S. Berg, N. J. Smith, R. Kern, M. Picus, S. Hoyer, M. H. van Kerkwijk, M. Brett, A. Haldane, J. F. del Río, M. Wiebe, P. Peterson, P. Gérard-Marchant, K. Sheppard, T....

    2020 doi

  4. [12]

    pfedbea: Combatting data heterogeneity for personalized federated learning by body exchange and aggregation abandon,

    J. He, D. Liu, S. Zhang, S. Ge, Y . Cao, and H. Tang, “pfedbea: Combatting data heterogeneity for personalized federated learning by body exchange and aggregation abandon,” in 2024 International Joint Conference on Neural Networks (IJCNN) , 2024, pp. 1–8

    2024

  5. [13]

    Deep residual learning for image recognition,

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

    2016

  6. [14]

    Measuring the effects of non- identical data distribution for federated visual classification,

    T.-M. H. Hsu, H. Qi, and M. Brown, “Measuring the effects of non- identical data distribution for federated visual classification,” arXiv preprint arXiv:1909.06335, 2019

    1909 arXiv

  7. [15]

    Stochastic controlled averaging for federated learning with communication compression,

    X. Huang, P. Li, and X. Li, “Stochastic controlled averaging for federated learning with communication compression,” in The Twelfth International Conference on Learning Representations , 2024

    2024

  8. [16]

    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

  9. [17]

    Matplotlib: A 2d graphics environment,

    J. D. Hunter, “Matplotlib: A 2d graphics environment,” Computing in Science & Engineering , vol. 9, no. 3, pp. 90–95, 2007

    2007

  10. [18]

    P. T. Inc. (2015) Collaborative data science. Montreal, QC. [Online]. Available: https://plot.ly

    2015

  11. [19]

    Heterogeneous personalized federated learning by local-global updates mixing via convergence rate,

    M. Jiang, A. Le, X. Li, and Q. Dou, “Heterogeneous personalized federated learning by local-global updates mixing via convergence rate,” in The Twelfth International Conference on Learning Representations , 2024

    2024

  12. [20]

    DepthFL : Depthwise federated learning for heterogeneous clients,

    M. Kim, S. Yu, S. Kim, and S.-M. Moon, “DepthFL : Depthwise federated learning for heterogeneous clients,” in The Eleventh International Conference on Learning Representations , 2023

    2023

  13. [21]

    Learning multiple layers of features from tiny images,

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

    2009

  14. [22]

    Lapan, Deep Reinforcement Learning Hands-On: Apply modern RL methods to practical problems of chatbots, robotics, discrete optimization, web automation, and more

    M. Lapan, Deep Reinforcement Learning Hands-On: Apply modern RL methods to practical problems of chatbots, robotics, discrete optimization, web automation, and more . Packt Publishing Ltd, 2020

    2020

  15. [23]

    A survey on federated learning systems: Vision, hype and reality for data privacy and protection,

    Q. Li, Z. Wen, Z. Wu, S. Hu, N. Wang, Y . Li, X. Liu, and B. He, “A survey on federated learning systems: Vision, hype and reality for data privacy and protection,” IEEE Transactions on Knowledge and Data Engineering, vol. 35, no. 4, pp. 3347–3366, 2021

    2021

  16. [24]

    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

  17. [25]

    Fedcompass: Efficient cross-silo federated learning on heterogeneous client devices using a computing power-aware scheduler,

    Z. Li, P. Chaturvedi, S. He, H. Chen, G. Singh, V . Kindratenko, E. A. Huerta, K. Kim, and R. Madduri, “Fedcompass: Efficient cross-silo federated learning on heterogeneous client devices using a computing power-aware scheduler,” in The Twelfth International Conference on Lear...

    2024

  18. [26]

    Federated learning with data- free distillation for heterogeneity-aware autonomous driving,

    J. Liang, J. Li, J. Zhang, and T. Zang, “Federated learning with data- free distillation for heterogeneity-aware autonomous driving,” in 2024 International Joint Conference on Neural Networks (IJCNN) , 2024, pp. 1–7

    2024

  19. [27]

    Continuous control with deep reinforcement learning,

    T. P. Lillicrap, J. J. Hunt, A. Pritzel, N. Heess, T. Erez, Y . Tassa, D. Silver, and D. Wierstra, “Continuous control with deep reinforcement learning,” in 4th International Conference on Learning Representations, ICLR 2016, San Juan, Puerto Rico, May 2-4, 2016, Conference Tr...

    2016

  20. [28]

    Towards distributed backdoor attacks with network detection in decentralized federated learning,

    B. Liu, Y . Xiao, R. Ye, Z. Ling, X. Ma, and B. Hui, “Towards distributed backdoor attacks with network detection in decentralized federated learning,” arXiv preprint arXiv:2501.15005 , 2025

    2025 arXiv

  21. [29]

    Integrating planning and deep reinforcement learning via automatic induction of task substructures,

    J.-C. Liu, C.-H. Chang, S.-H. Sun, and T.-L. Yu, “Integrating planning and deep reinforcement learning via automatic induction of task substructures,” in The Twelfth International Conference on Learning Representations , 2024

    2024

  22. [30]

    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

  23. [31]

    Pytorch: An imperative style, high-performance deep learning library,

    A. Paszke, S. Gross, F. Massa, A. Lerer, J. Bradbury, G. Chanan, T. Killeen, Z. Lin, N. Gimelshein, L. Antiga et al. , “Pytorch: An imperative style, high-performance deep learning library,” Advances in neural information processing systems , vol. 32, 2019

    2019

  24. [32]

    Plaat, Deep reinforcement learning

    A. Plaat, Deep reinforcement learning . Springer, 2022, vol. 10

    2022

  25. [33]

    PeFLL: Personalized federated learning by learning to learn,

    J. Scott, H. Zakerinia, and C. H. Lampert, “PeFLL: Personalized federated learning by learning to learn,” in The Twelfth International Conference on Learning Representations , 2024

    2024

  26. [34]

    Reinforcement learning: An introduction,

    R. S. Sutton, “Reinforcement learning: An introduction,” A Bradford Book, 2018

    2018

  27. [35]

    Fedimpro: Measuring and improving client update in federated learning,

    Z. Tang, Y . Zhang, S. Shi, X. Tian, T. Liu, B. Han, and X. Chu, “Fedimpro: Measuring and improving client update in federated learning,” in The Twelfth International Conference on Learning Representations , 2024

    2024

  28. [36]

    SciPy 1.0: Fundamental Algorithms for Scientific Computing in Python,

    P. Virtanen, R. Gommers, T. E. Oliphant, M. Haberland, T. Reddy, D. Cournapeau, E. Burovski, P. Peterson, W. Weckesser, J. Bright, S. J. van der Walt, M. Brett, J. Wilson, K. J. Millman, N. Mayorov, A. R. J. Nelson, E. Jones, R. Kern, E. Larson, C. J. Carey, ˙I. Polat, Y . Fen...

    2020

  29. [37]

    Confidence region estimation techniques for nonlinear regression in groundwater flow: Three case studies,

    K. W. Vugrin, L. P. Swiler, R. M. Roberts, N. J. Stucky-Mack, and S. P. Sullivan, “Confidence region estimation techniques for nonlinear regression in groundwater flow: Three case studies,” Water Resources Research, vol. 43, no. 3, 2007

    2007

  30. [38]

    FedCDA: Federated learning with cross-rounds divergence-aware aggregation,

    H. Wang, H. Xu, Y . Li, Y . Xu, R. Li, and T. Zhang, “FedCDA: Federated learning with cross-rounds divergence-aware aggregation,” in The Twelfth International Conference on Learning Representations , 2024

    2024

  31. [39]

    Tackling the data heterogeneity in asynchronous federated learning with cached update calibration,

    Y . Wang, Y . Cao, J. Wu, R. Chen, and J. Chen, “Tackling the data heterogeneity in asynchronous federated learning with cached update calibration,” in The Twelfth International Conference on Learning Representations, 2024

    2024

  32. [40]

    Fedhyper: A universal and robust learning rate scheduler for federated learning with hypergradient descent,

    Z. Wang, J. Wang, and A. Li, “Fedhyper: A universal and robust learning rate scheduler for federated learning with hypergradient descent,” in The Twelfth International Conference on Learning Representations , 2024

    2024

  33. [41]

    Fedinverse: Evaluating privacy leakage in federated learning,

    D. Wu, J. Bai, Y . Song, J. Chen, W. Zhou, Y . Xiang, and A. Sajjanhar, “Fedinverse: Evaluating privacy leakage in federated learning,” in The Twelfth International Conference on Learning Representations, ICLR 2024, Vienna, Austria, May 7-11, 2024 . OpenReview.net, 2024

    2024

  34. [42]

    Personalized federated learning for intelligent iot applications: A cloud-edge based framework,

    Q. Wu, K. He, and X. Chen, “Personalized federated learning for intelligent iot applications: A cloud-edge based framework,” IEEE Open Journal of the Computer Society , vol. 1, pp. 35–44, 2020

    2020

  35. [43]

    H. Xiao, K. Rasul, and R. V ollgraf. (2017) Fashion-mnist: a novel image dataset for benchmarking machine learning algorithms

    2017

  36. [44]

    Trustworthy personalized bayesian federated learning via posterior fine-tune,

    C. Xu, M. Luo, and E. E. Kuruoglu, “Trustworthy personalized bayesian federated learning via posterior fine-tune,” in 2024 International Joint Conference on Neural Networks (IJCNN) , 2024, pp. 1–8

    2024

  37. [45]

    Byzantine-robust distributed learning: Towards optimal statistical rates,

    D. Yin, Y . Chen, R. Kannan, and P. Bartlett, “Byzantine-robust distributed learning: Towards optimal statistical rates,” in International conference on machine learning . Pmlr, 2018, pp. 5650–5659

    2018

  38. [46]

    Exploiting class feature alignment for personalized federated learning in mixed skew scenarios,

    W. Yin and Y . Zhang, “Exploiting class feature alignment for personalized federated learning in mixed skew scenarios,” in 2024 International Joint Conference on Neural Networks (IJCNN) , 2024, pp. 1–8

    2024

  39. [47]

    Turning the curse of heterogeneity in federated learning into a blessing for out-of-distribution detection,

    S. Yu, J. Hong, H. Wang, Z. Wang, and J. Zhou, “Turning the curse of heterogeneity in federated learning into a blessing for out-of-distribution detection,” in The Eleventh International Conference on Learning Representations, 2023

    2023

  40. [48]

    Bayesian nonparametric federated learning of neural networks,

    M. Yurochkin, M. Agarwal, S. Ghosh, K. Greenewald, N. Hoang, and Y . Khazaeni, “Bayesian nonparametric federated learning of neural networks,” in International conference on machine learning . PMLR, 2019, pp. 7252–7261

    2019

  41. [49]

    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, 2023, pp. 11 237–11 244

    2023

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.