REVIEW 3 major objections 4 minor 39 references
Incentivizing Truthful Collaboration in Heterogeneous Federated Learning
T0 review · 3 major / 4 minor · reviewed 2026-08-12 · deepseek-v4-flash
Pith's one-line read In heterogeneous federated learning, a budget-balanced payment rule can make truthful gradient reporting a near best response.
desk verdict A good mechanism-design idea with a load-bearing proof gap: the stated BIC bound in Theorem 5.1 does not follow from the appendix's per-step estimates, and Theorem 5.2 has a similar mismatch. read the letter →
The pith
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
The reading
What carries the argument
The load-bearing object is the squared-norm-difference payment rule together with a trajectory-divergence bound. Claim E.1 shows that with all other clients truthful, two trajectories produced by different strategies of client $i$ remain close: the expected squared distance contracts by a factor $c_t = 2(1-2\gamma_t m + \gamma_t^2 H^2)$ per step, up to a term proportional to the client's deviation from truthful scaling and injected noise. Claim E.2 turns that closeness into a bound on any reward difference the client could gain by steering, and Claim E.3 shows the payment grows with the same deviation. Setting the payment constant so the accumulated penalty coefficient $\bar{C}_t = \prod_{l=t+1}^T c_l$ dominates the accumulated reward gain makes truthfulness a near-best response.
What would settle it
Take a simple two-client strongly convex quadratic problem with known $m$, $H$, and $L$, run FedSGD with the theorem's payment constant, and measure client 1's utility as its scaling factor ranges over $[1,3]$ while client 2 is truthful. If the utility is not maximized at scaling factor $1$ within the promised $\varepsilon$, or if the measured best response deviates by more than $\varepsilon$ from truth, the central claim fails.
Extended reading notes
Core claim
The central claim is that the incentive to manipulate gradients in heterogeneous federated learning is a steering effect—a client can pull the server model toward its own data distribution by sending an amplified update—and that this steering can be neutralized by comparing each report against the others. The paper's payment at step $t$ is $p_i^t = C_t\left(\lVert m_i^t\rVert^2 - \frac{1}{N-1}\sum_{j\neq i}\lVert m_j^t\rVert^2\right)$, which sums to zero over clients, so the server neither profits nor loses. Theorem 5.1 states that when each client objective $F_i$ is $H$-smooth and $m$-strongly-convex, each reward $R_i$ is $L$-Lipschitz, and the payment constant is chosen as $\sqrt{2\bar{C}_t\gamma_t L}/(N\varepsilon)$, the FedSGD protocol is $\sqrt{2LG\varepsilon/N}$-Bayesian incentive compatible, and the best response of any client who believes others are truthful is $\varepsilon$-approximately truthful. The proof bounds the divergence between a manipulated and a truthful trajectory, then shows the payment dominates any reward gain that divergence could buy.
Load-bearing premise
The load-bearing premise is that all client objectives are uniformly $m$-strongly convex and $H$-smooth, rewards are $L$-Lipschitz, and these constants are known well enough for the server to set the payment scale; the experiments replace the theoretical constant with hand-tuned values because real neural-network objectives are non-convex and the constants are unknown.
Editorial extensions
If this is right
- A unilateral deviation from truthful reporting buys any client at most an $\varepsilon$ utility gain when all other clients report truthfully.
- The protocol can be analyzed and run as if updates were honest: approximately truthful best responses keep the FedSGD convergence rate $O(1/(NT))$ for fixed heterogeneity bounds, and $O(1/(\sqrt{N}T))$ if heterogeneity grows as $N^{1/4}$.
- Because the payments sum to zero, the mechanism is budget-balanced; the server has no financial stake and no extra communication round is required to compute the penalties.
- Total payment per client is bounded by $O(1/N)$ in the constant-heterogeneity regime, so the incentive layer does not impose a growing per-client cost as the federation scales.
- The same payment rule is shown empirically to flatten the utility curve of a misreporting client on FeMNIST, Shakespeare, and Twitter under FedSGD, median-based FedSGD, and FedAvg, indicating the mechanism transfers to non-convex models.
Reading between the lines
- The payment is a relative norm comparison, so if all clients scaled their updates by the same factor, the penalty terms would cancel; this suggests the mechanism is aimed at unilateral deviation, and coalitional scaling by a group of clients is a natural stress test not analyzed in the paper.
- The theorem's payment constant depends on $m$, $H$, and $L$, which are unknown in practice; the experiments hand-tune the analogous constant per dataset, so a plug-in estimator or adaptive tuning scheme would be needed to make the guarantee operational.
- Because the penalty grows with squared norms, the rule may also act as a defense against outlier updates, which would connect it to outlier-resistant aggregation; the paper only demonstrates this empirically via median aggregation, leaving the theoretical link open.
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The paper studies incentives for gradient manipulation in heterogeneous federated learning. After giving examples where a client benefits from upscaling its gradient, it defines a game in which clients can scale their updates and add noise, and proposes a budget-balanced payment rule (Section 5, Eq. (1)) that charges a client the squared norm of its message minus the average squared norm of the other messages. The main theoretical claims are that, under smooth strongly convex objectives and Lipschitz rewards, FedSGD with this payment rule is sqrt(2 L G epsilon / N)-Bayesian incentive compatible and induces epsilon-approximately truthful reporting (Theorem 5.1), with explicit payment and convergence bounds (Theorems 5.2 and 5.3). The paper also reports experiments on FeMNIST, Twitter, and Shakespeare with FedSGD, median-based FedSGD, and FedAvg.
Significance. If the main theorem were fully established, the paper would make a useful contribution: it would show that a simple, communication-free, budget-balanced payment rule can restore approximate truthfulness under heterogeneity, using only a magnitude-based heterogeneity assumption rather than a direction-based one. The per-step trajectory and utility decomposition in Appendix E is a reasonable proof strategy, and the experimental section covers several tasks and aggregation rules. However, the global BIC bound and the payment bound do not follow from the supplied proofs as printed, so the paper's central theoretical contribution is presently unsupported. The empirical results are suggestive but are obtained with hand-tuned penalty magnitudes on non-convex models, so they do not validate the quantitative theoretical guarantees.
major comments (3)
- [Appendix E, Proposition E.4 and Claim E.5; Theorem 5.1] The transition from per-step utility bounds to the claimed total BIC bound is not justified. Proposition E.4 bounds the utility gain of a one-step deviation by sqrt(2 C_t gamma_t L epsilon)/N, where C_t is the backward product; summing over t gives a total of order (sqrt(2 L epsilon)/N) * sum_t sqrt(C_t gamma_t). Claim E.5 and Theorem 5.1 instead assert a bound of sqrt(2 L G epsilon)/N with G = sum_t gamma_t sqrt(C_t), which is proportional to sqrt(sum_t gamma_t sqrt(C_t)). The inequality sum_t sqrt(C_t gamma_t) <= sqrt(sum_t gamma_t sqrt(C_t)) needed for this step is false; for T=2, gamma_1=gamma_2=1, C_1=4, C_2=1, the left side is 3 and the right side is sqrt(3). Thus Theorem 5.1(1) is not established as written.
- [Theorem 5.2 and Appendix F] The stated payment bound does not follow from the appendix's per-step bound. The proof gives p_i^t/C_t <= 2 epsilon^2 + 2 epsilon sigma + 2 zeta^2 + rho^2 + 2 epsilon ||nabla F_i(theta_t)||, with C_t = sqrt(2 C_t gamma_t L)/(N epsilon). Substituting and summing yields a term proportional to (sqrt(2L)/(N epsilon)) * (2 epsilon^2 + 2 epsilon sigma + 2 zeta^2 + rho^2) * sum_t sqrt(C_t gamma_t), which contains a (2 zeta^2 + rho^2)/epsilon divergence and a sum over sqrt(C_t gamma_t). The printed bound in Theorem 5.2 has no 1/epsilon term and instead uses sqrt(2 L G)/N with G = sum_t gamma_t sqrt(C_t). These expressions are not equivalent, so Theorem 5.2 is unsupported by Appendix F.
- [Section 5.1 and Section 6.1] The theoretical payment coefficient C_t in Theorem 5.1 depends on m, H, L, epsilon, and the whole future learning-rate schedule through the backward product C_t, yet the experiments use non-convex neural networks with per-dataset hand-tuned penalty constants C (Section 6.1, Figures 2-5). No estimation procedure for the constants m, H, L is provided, so the theorem's prescribed payment value cannot be evaluated in the demonstrated setting. The paper should state this limitation explicitly and, ideally, discuss how the constants can be estimated or which robustness properties hold when they are misspecified.
minor comments (4)
- [Theorem 5.1] The same symbol C_t is used both for the payment coefficient and for the backward product C_t = prod_{l=t+1}^T c_l in the same statement; please introduce distinct notation to remove the ambiguity.
- [Appendix E, Claim E.2] Claim E.2 defines C_t as a forward product prod_{t'=1}^t c_{t'}, while Corollary 1 and Proposition E.4 use the backward product prod_{l=t+1}^T c_l; this inconsistency should be corrected so the telescoping argument is readable.
- [Section 6.1] The experimental utility is defined as -F_i(theta_final) - C * sum_t p_i^t, but the relation between the experimental penalty magnitude C and the theoretical coefficient C_t from Eq. (1) is not specified; the plots therefore cannot be compared with the quantitative bounds of Theorems 5.1-5.3.
- [Appendix G, alternative bound] In the statement of the alternative bound using Chung's lemma, the constant c1 is written as 16H(epsilon^2 + M)/(N m^2), omitting the M_V zeta^2 term that appears in the main proof; this looks like a typo and should be fixed.
Circularity Check
No circularity; the incentive-compatibility theorem is derived from stated assumptions and a deliberately constructed payment rule, with no fitted input or load-bearing self-citation.
full rationale
The paper's central claim, Theorem 5.1, is a mechanism-design result: given H-smooth, m-strongly-convex client objectives and L-Lipschitz rewards, the quadratic payment rule is shown to make truthful reporting epsilon-close to optimal and the best response epsilon-approximately truthful. The proof proceeds from the stated smoothness, strong-convexity, Lipschitz, and bounded-heterogeneity assumptions, using standard external lemmas (e.g., Bottou et al. 2018, Chung 1954). No parameter is fitted to data: epsilon is an explicit design target, and the payment constant is expressed in terms of assumed known constants. The fact that the payment rule is deliberately quadratic in reported gradient norms is the mechanism itself, not a hidden equivalence: the theorem's content is that with the stated choice of constant the quadratic penalty dominates any Lipschitz-bounded reward gain. This is normal mechanism design, not circularity. The self-citations in the paper (e.g., Tsoy et al. 2024 and Dorner et al. 2023, which share authors with the present work) appear only as background motivation in the introduction and related work; the theoretical results do not import their conclusions, and no uniqueness theorem or ansatz from the authors' prior work is load-bearing. The experiments hand-tune the payment scale and the paper explicitly acknowledges this departure from the theoretical C_t, which is a scope limitation rather than a fitted input renamed as a prediction. The concerns raised by the skeptic about Appendix E's summation step, the overloaded C_t notation in Theorem 5.1, and the C_t substitution in Theorem 5.2 are substantive correctness and well-definedness issues that may invalidate the bounds as written, but they do not make the claimed theorem equivalent to its inputs or to a fitted quantity. Accordingly, no circular step can be exhibited, and the appropriate circularity score is 0.
Assumptions & free parameters
free parameters (2)
- Experimental payment magnitude C =
grid-searched per task: values from 1e-6 to 2e-3
- Approximation tolerance epsilon =
not set in experiments
assumptions (7)
- domain assumption Each client's objective F_i is m-strongly convex and H-smooth, and each sample loss f(theta; z) is convex and smooth.
- domain assumption Stochastic gradients are unbiased estimators of full gradients with variance bounded by sigma^2.
- ad hoc to paper Bounded magnitude-only gradient heterogeneity: | ||grad F_i(theta)||^2 - ||grad F(theta)||^2 | <= zeta^2.
- ad hoc to paper Bounded variance difference: | E||e_i(theta)||^2 - E||e_j(theta)||^2 | <= rho^2.
- domain assumption Reward R_i(theta) is L-Lipschitz in theta.
- domain assumption Client actions are restricted to scaling factors |a_i^t| >= 1 and noise levels b_i^t >= 0.
- domain assumption Gradient error variance is bounded as E||e_i(theta_t)||^2 <= M + M_V ||grad F_i(theta_t)||^2.
Cite this review
Pith. "Pith review of Incentivizing Truthful Collaboration in Heterogeneous Federated Learning." pith.science (2026). https://pith.science/paper/HLUNIYC5
@misc{pith2026241200980,
author = {Pith},
title = {Pith review of: Incentivizing Truthful Collaboration in Heterogeneous Federated Learning},
year = {2026},
howpublished = {\url{https://pith.science/paper/HLUNIYC5}},
note = {Machine review of arXiv:2412.00980}
}
read the original abstract
Federated learning (FL) is a distributed collaborative learning method, where multiple clients learn together by sharing gradient updates instead of raw data. However, it is well-known that FL is vulnerable to manipulated updates from clients. In this work we study the impact of data heterogeneity on clients' incentives to manipulate their updates. First, we present heterogeneous collaborative learning scenarios where a client can modify their updates to be better off, and show that these manipulations can lead to diminishing model performance. To prevent such modifications, we formulate a game in which clients may misreport their gradient updates in order to "steer" the server model to their advantage. We develop a payment rule that provably disincentivizes sending modified updates under the FedSGD protocol. We derive explicit bounds on the clients' payments and the convergence rate of the global model, which allows us to study the trade-off between heterogeneity, payments and convergence. Finally, we provide an experimental evaluation of the effectiveness of our payment rule in the FedSGD, median-based aggregation FedSGD and FedAvg protocols on three tasks in computer vision and natural language processing. In all cases we find that our scheme successfully disincentivizes modifications.
Figures
Figures from the paper (5 more)
Reference graph
Works this paper leans on
-
[1]
Mitigating bias in federated learning
Annie Abay, Yi Zhou, Nathalie Baracaldo, Shashank Rajamoni, Ebube Chuba, and Heiko Ludwig. Mitigating bias in federated learning. arXiv preprint arXiv:2012.02447, 2020
arXiv 2012
-
[2]
Byzantine stochastic gradient descent
Dan Alistarh, Zeyuan Allen-Zhu, and Jerry Li. Byzantine stochastic gradient descent. Conference on Neural Information Processing Systems (NeurIPS), 2018
work page 2018
-
[3]
Machine learning with adversaries: Byzantine tolerant gradient descent
Peva Blanchard, El Mahdi El Mhamdi, Rachid Guerraoui, and Julien Stainer. Machine learning with adversaries: Byzantine tolerant gradient descent. Conference on Neural Information Processing Systems (NIPS), 2017
work page 2017
-
[4]
One for one, or all for all: Equilibria and optimality of collaboration in federated learning
Avrim Blum, Nika Haghtalab, Richard Lanas Phillips, and Han Shao. One for one, or all for all: Equilibria and optimality of collaboration in federated learning. In International Conference on Machine Learning, pages 1005--1014. PMLR, 2021
work page 2021
-
[5]
Optimization methods for large-scale machine learning
L \'e on Bottou, Frank E Curtis, and Jorge Nocedal. Optimization methods for large-scale machine learning. SIAM review, 60 0 (2): 0 223--311, 2018
2018
-
[6]
Leaf: A benchmark for federated settings
Sebastian Caldas, Sai Meher Karthik Duddu, Peter Wu, Tian Li, Jakub Kone c n \`y , H Brendan McMahan, Virginia Smith, and Ameet Talwalkar. Leaf: A benchmark for federated settings. arXiv preprint arXiv:1812.01097, 2018
arXiv 2018
-
[7]
Linear speedup in personalized collaborative learning
El Mahdi Chayti, Sai Praneeth Karimireddy, Sebastian U Stich, Nicolas Flammarion, and Martin Jaggi. Linear speedup in personalized collaborative learning. arXiv preprint arXiv:2111.05968, 2021
arXiv 2021
-
[8]
On a stochastic approximation method
Kai Lai Chung. On a stochastic approximation method. The Annals of Mathematical Statistics, pages 463--483, 1954
work page 1954
Show all 39 references
-
[9]
Model-sharing games: Analyzing federated learning under voluntary participation
Kate Donahue and Jon Kleinberg. Model-sharing games: Analyzing federated learning under voluntary participation. AAAI Conference on Artificial Intelligence, 2021 a
2021
-
[10]
Optimality and stability in federated learning: A game-theoretic approach
Kate Donahue and Jon Kleinberg. Optimality and stability in federated learning: A game-theoretic approach. Conference on Neural Information Processing Systems (NeurIPS), 2021 b
2021
-
[11]
Incentivizing honesty among competitors in collaborative learning and optimization
Florian E Dorner, Nikola Konstantinov, Georgi Pashaliev, and Martin Vechev. Incentivizing honesty among competitors in collaborative learning and optimization. Conference on Neural Information Processing Systems (NeurIPS), 2023
2023
-
[12]
The role of cross-silo federated learning in facilitating data sharing in the agri-food sector
Aiden Durrant, Milan Markovic, David Matthews, David May, Jessica Enright, and Georgios Leontidis. The role of cross-silo federated learning in facilitating data sharing in the agri-food sector. Computers and Electronics in Agriculture, 193: 0 106648, 2022
2022
-
[13]
Robust federated learning with noisy and heterogeneous clients
Xiuwen Fang and Mang Ye. Robust federated learning with noisy and heterogeneous clients. In Conference on Computer Vision and Pattern Recognition (CVPR), 2022
2022
-
[14]
Application of logistic function for analysis of marginal value diminishing laws
Stasys Girdzijauskas and Dalia Štreimikienė. Application of logistic function for analysis of marginal value diminishing laws. Ekonomika, 78: 0 34–47, Dec. 2007. doi:10.15388/Ekon.2007.17611. URL https://www.journals.vu.lt/ekonomika/article/view/17611
2007
-
[15]
Sharp bounds for federated averaging (local sgd) and continuous perspective
Margalit R Glasgow, Honglin Yuan, and Tengyu Ma. Sharp bounds for federated averaging (local sgd) and continuous perspective. In International Conference on Artificial Intelligence and Statistics, pages 9050--9090. PMLR, 2022
2022
-
[16]
On the effect of defections in federated learning and how to prevent them
Minbiao Han, Kumar Kshitij Patel, Han Shao, and Lingxiao Wang. On the effect of defections in federated learning and how to prevent them. arXiv preprint arXiv:2311.16459, 2023
2023 arXiv
-
[17]
Evaluating and incentivizing diverse data contributions in collaborative learning
Baihe Huang, Sai Praneeth Karimireddy, and Michael I Jordan. Evaluating and incentivizing diverse data contributions in collaborative learning. arXiv preprint arXiv:2306.05592, 2023
2023 arXiv
-
[18]
Advances and open problems in federated learning
Peter Kairouz, H Brendan McMahan, Brendan Avent, Aur \'e lien Bellet, Mehdi Bennis, Arjun Nitin Bhagoji, Kallista Bonawitz, Zachary Charles, Graham Cormode, Rachel Cummings, et al. Advances and open problems in federated learning. Foundations and Trends in Machine Learning , 2021
2021
-
[19]
Scaffold: Stochastic controlled averaging for federated learning
Sai Praneeth Karimireddy, Satyen Kale, Mehryar Mohri, Sashank Reddi, Sebastian Stich, and Ananda Theertha Suresh. Scaffold: Stochastic controlled averaging for federated learning. In International conference on machine learning, pages 5132--5143. PMLR, 2020
2020
-
[20]
Mechanisms that incentivize data sharing in federated learning
Sai Praneeth Karimireddy, Wenshuo Guo, and Michael Jordan. Mechanisms that incentivize data sharing in federated learning. In Workshop on Federated Learning: Recent Advances and New Challenges (in Conjunction with NeurIPS 2022), 2022
2022
-
[21]
Tighter theory for local sgd on identical and heterogeneous data
Ahmed Khaled, Konstantin Mishchenko, and Peter Richt \'a rik. Tighter theory for local sgd on identical and heterogeneous data. In International Conference on Artificial Intelligence and Statistics, pages 4519--4529. PMLR, 2020
2020
-
[22]
A unified theory of decentralized sgd with changing topology and local updates
Anastasia Koloskova, Nicolas Loizou, Sadra Boreiri, Martin Jaggi, and Sebastian Stich. A unified theory of decentralized sgd with changing topology and local updates. In International Conference on Machine Learning, pages 5381--5393. PMLR, 2020
2020
-
[23]
Federated learning for open banking
Guodong Long, Yue Tan, Jing Jiang, and Chengqi Zhang. Federated learning for open banking. In Federated learning: privacy and incentive, pages 240--254. Springer, 2020
2020
-
[24]
Three approaches for personalization with applications to federated learning
Yishay Mansour, Mehryar Mohri, Jae Ro, and Ananda Theertha Suresh. Three approaches for personalization with applications to federated learning. arXiv preprint arXiv:2002.10619, 2020
2002 arXiv
-
[25]
Personalized federated learning through local memorization
Othmane Marfoq, Giovanni Neglia, Richard Vidal, and Laetitia Kameni. Personalized federated learning through local memorization. In International Conference on Machine Learing (ICML), 2022
2022
-
[26]
Communication-efficient learning of deep networks from decentralized data
Brendan McMahan, Eider Moore, Daniel Ramage, Seth Hampson, and Blaise Aguera y Arcas. Communication-efficient learning of deep networks from decentralized data. In Conference on Uncertainty in Artificial Intelligence (AISTATS), 2017
2017
-
[27]
Partially personalized federated learning: Breaking the curse of data heterogeneity
Konstantin Mishchenko, Rustem Islamov, Eduard Gorbunov, and Samuel Horv \'a th. Partially personalized federated learning: Breaking the curse of data heterogeneity. arXiv preprint arXiv:2305.18285, 2023
2023 arXiv
-
[28]
Algorithmic Game Theory
Noam Nisan, Eva Tardos, Tim Roughgarden, and Vijay Vazirani. Algorithmic Game Theory. Cambridge University Press, 2007
2007
-
[29]
Federated learning techniques applied to credit risk management: A systematic literature review
Adil Oualid, Yassine Maleh, and Lahcen Moumoun. Federated learning techniques applied to credit risk management: A systematic literature review. EDPACS, 68 0 (1): 0 42--56, 2023
2023
-
[30]
The limits and potentials of local sgd for distributed heterogeneous learning with intermittent communication
Kumar Kshitij Patel, Margalit Glasgow, Ali Zindari, Lingxiao Wang, Sebastian U Stich, Ziheng Cheng, Nirmit Joshi, and Nathan Srebro. The limits and potentials of local sgd for distributed heterogeneous learning with intermittent communication. Conference on Learning Theory (CO...
2024
-
[31]
Robust aggregation for federated learning
Krishna Pillutla, Sham M Kakade, and Zaid Harchaoui. Robust aggregation for federated learning. IEEE Transactions on Signal Processing, 70: 0 1142--1154, 2022
2022
-
[32]
The future of digital health with federated learning
Nicola Rieke, Jonny Hancox, Wenqi Li, Fausto Milletari, Holger R Roth, Shadi Albarqouni, Spyridon Bakas, Mathieu N Galtier, Bennett A Landman, Klaus Maier-Hein, et al. The future of digital health with federated learning. NPJ digital medicine, 3 0 (1): 0 1--7, 2020
2020
-
[33]
Back to the drawing board: A critical evaluation of poisoning attacks on production federated learning
Virat Shejwalkar, Amir Houmansadr, Peter Kairouz, and Daniel Ramage. Back to the drawing board: A critical evaluation of poisoning attacks on production federated learning. In IEEE Symposium on Security and Privacy, 2022
2022
-
[34]
Federated machine learning in vehicular networks: A summary of recent applications
Kang Tan, Duncan Bremner, Julien Le Kernec, and Muhammad Imran. Federated machine learning in vehicular networks: A summary of recent applications. In 2020 international conference on UK-China emerging technologies (UCET), pages 1--4. IEEE, 2020
2020
-
[35]
Provable mutual benefits from federated learning in privacy-sensitive domains
Nikita Tsoy, Anna Mihalkova, Teodora N Todorova, and Nikola Konstantinov. Provable mutual benefits from federated learning in privacy-sensitive domains. In Conference on Uncertainty in Artificial Intelligence (AISTATS), 2024
2024
-
[36]
Incentive mechanisms for federated learning: From economic and game theoretic perspective
Xuezhen Tu, Kun Zhu, Nguyen Cong Luong, Dusit Niyato, Yang Zhang, and Juan Li. Incentive mechanisms for federated learning: From economic and game theoretic perspective. arXiv preprint arXiv:2111.11850, 2021
2021 arXiv
-
[37]
Minibatch vs local sgd for heterogeneous distributed learning
Blake E Woodworth, Kumar Kshitij Patel, and Nati Srebro. Minibatch vs local sgd for heterogeneous distributed learning. Advances in Neural Information Processing Systems, 33: 0 6281--6292, 2020
2020
-
[38]
Byzantine-robust distributed learning: Towards optimal statistical rates
Dong Yin, Yudong Chen, Ramchandran Kannan, and Peter Bartlett. Byzantine-robust distributed learning: Towards optimal statistical rates. In International Conference on Machine Learning, 2018
2018
-
[39]
A survey of incentive mechanism design for federated learning
Yufeng Zhan, Jie Zhang, Zicong Hong, Leijie Wu, Peng Li, and Song Guo. A survey of incentive mechanism design for federated learning. IEEE Transactions on Emerging Topics in Computing, 10 0 (2): 0 1035--1044, 2021
2021
Reviewed August 12, 2026 · model on record in the stance chip above.
Discussion (0). Continue with ORCID to comment.