Pith. sign in

REVIEW 4 major objections 6 minor 32 references

Enhancing Federated Survival Analysis through Peer-Driven Client Reputation in Healthcare

T0 review · 4 major / 6 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 federated survival-analysis framework uses differentially privatized peer feedback to build client reputation, down-weighting noisy hospitals and preserving C-index accuracy.

desk verdict A plausible reputation-based FL/Cox framework whose central mechanism is unproven: the feedback signal is assumed to track true contribution quality, and the convergence proof in Section 4.5 is algebraically off. read the letter →

arxiv 2505.16190 v1 pith:QA7TVAJY submitted 2025-05-22 cs.LG

Navid Seidi , Satyaki Roy , Sajal Das This is my paper · ORCID
classification cs.LG
keywords FederatedLearningSurvivalAnalysisCoxProportionalHazardsDecentralizedAggregationHealthcareInformaticsDataHeterogeneityPrivacy-PreservingMachineConcordanceIndex
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 survival analysis in healthcare can be made trustworthy without a central reputation authority: each client rates a peer by how much that peer's model update improves the concordance index on the client's own local data, and these ratings accumulate over rounds into reputation scores that weight the aggregation of Cox proportional-hazards models. The reputation channel is deliberately separated from the learning channel, with updates differentially privatized before they are shown to peers and unaltered updates sent to the server, so privacy protection does not degrade the global model. A clustering step groups institutions by feature completeness and risk profile to handle data heterogeneity and to filter noisy feedback. If the claim holds, hospitals can collaboratively fit survival models whose quality degrades gracefully when some participants gradually begin poisoning their updates, and no node needs to see another node's data. Reported experiments on synthetic data and the SEER breast-cancer registry show stable, high concordance-index values and clear down-weighting of noisy clients relative to reputation-free federated averaging.

What carries the argument

The load-bearing object is the peer-feedback message $m_{j,k}(t)=\Omega(M(\theta_j(t),\theta_k(t)))-\Omega(M(\theta_j(t)))$, in which client $j$ scores client $k$'s update by the change in concordance index $\Omega$ of a Cox model evaluated on $j$'s local data with and without $k$'s update. This message, weighted by the sender's reputation $RS_{ij}(t)$, drives the update $RS_{ik}(t+1)=RS_{ik}(t)+\alpha\sum_j RS_{ij}(t)m_{j,k}(t)$, and the resulting scores set client-selection probabilities $P(i)=RS_i/\sum_{j\in C_k}RS_j$ inside each cluster. The privacy half of the mechanism clips updates to norm $Q$ and adds Gaussian noise $\sigma=Q\sqrt{2\ln(1.25/\delta)}/\zeta$ before peer evaluation, while unperturbed updates flow to the server. Clustering by feature-completeness vectors and predicted risk, with a concordance term weighted by $\lambda$, turns peer evaluations into cluster-specific, noise-resistant aggregation.

What would settle it

Run the scheme under heterogeneous local datasets and compare each client's reputation trajectory with the effect of its update on a held-out global evaluation set across rounds; if clients with rising reputations consistently contribute nothing or negatively to global concordance while low-reputation clients contribute strongly, the peer-feedback signal is not faithful and the central benefit collapses. Equivalently, inject a fixed constant bias into the feedback messages $m_{j,k}$ and check whether the Section 4.6 error bounds still hold.

Watch

Extended reading notes

Core claim

Working with the Cox proportional-hazards model, the central claim is that a reputation score built from peer feedback, specifically the change in a peer's local concordance index when a client's update is added, can serve as a reliable aggregation weight in heterogeneous federated learning, as long as the feedback itself is reliable. The paper further claims that differential privacy can be applied to client updates before peer evaluation without materially changing the outcome: the expected errors induced by DP noise in reputation feedback, reputation scores, client-selection probabilities, and the aggregated global model are each bounded by explicit constants involving the privacy parameters. A formal lemma states that, under zero-mean feedback noise, estimated reputations converge to the true reliability of each peer over rounds. Empirically, the method keeps the global C-index near 0.66 across ten synthetic rounds, beats the no-reputation baseline consistently, and matches or exceeds the reputation-based TFFL baseline on most SEER states.

Load-bearing premise

The load-bearing premise is that when a client's update improves a peer's local concordance index, that update is genuinely helpful to the global survival model; if local gains diverge from global contribution, reputation weighting could reward exactly the wrong clients.

Editorial extensions

If this is right

  • A hospital that starts injecting noise after an honest warm-up sees its reputation decline over rounds, so its influence on the shared survival model shrinks.
  • Reputation can be computed without any central trusted dataset, because each evaluation uses only local concordance-index changes and privatized updates.
  • The shared Cox model can reach accuracy comparable to reputation-free federated averaging while keeping each client's update private from its peers.
  • Institutions with similar feature-completeness and risk profiles can be clustered to improve concordance under data heterogeneity.
  • Reputation updates can be throttled to cut message overhead, at the cost of some accuracy, so an intermediate update frequency balances communication and performance.

Reading between the lines

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

  • Looking beyond the paper, the same DP decoupling should transfer to other federated objectives: any secondary client score whose value changes smoothly with the update could be computed on a privatized channel while the clean channel serves the global objective.
  • The convergence lemma assumes zero-mean feedback noise; the paper's static-bias experiment suggests resilience but does not prove it, so a natural extension is to derive reputation-error bounds under fixed non-zero mean bias.
  • Whether the mechanism helps the global model ultimately hinges on local concordance gains tracking global contribution; a direct test is to correlate each client's reputation trajectory with its effect on a held-out global evaluation set after every round.
  • One could also apply the reputation-weighted, cluster-selection scheme to other survival losses, such as discrete-time or deep survival models, where the same concordance-based feedback and DP decoupling would carry over.
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

4 major / 6 minor

Summary. The paper proposes a federated survival analysis framework using Cox proportional hazards models with a peer-driven reputation mechanism. Reputation scores are updated from peer feedback defined as the change in a peer's local concordance index when incorporating another client's update. Clients are clustered by feature-completeness vectors, updates are DP-noised for reputation sharing, and global aggregation weights updates by reputation-based selection probabilities. Theoretical claims include convergence of reputation estimates to true reliability and DP accuracy guarantees; experiments on synthetic data and SEER compare against TFFL and no-reputation FedAvg.

Significance. If the mechanism were shown to make reputation track true contribution quality, the paper would address a real problem in federated healthcare. The decoupling of DP-noised peer feedback from clean server aggregation is an interesting architectural choice, and the SEER evaluation is a strength. However, the central convergence proof assumes the very property at issue, contains an algebraic error, and the experiments do not establish that local C-index improvements correlate with global model quality. The theoretical contribution is therefore not load-bearing, and the empirical support is insufficient to support the abstract's claims.

major comments (4)
  1. [Section 4.5, Lemma] The proof of the Lemma assumes m_wv(t) = RS*_v + eta, i.e., that the feedback signal is already the true reliability plus zero-mean noise. This is exactly the property the lemma is supposed to establish, so the convergence argument is circular. The paper itself notes this at the end of the proof ('provided the feedback mechanism is ideal'), which is a limitation statement but it is placed in a proof of a theorem, not as a caveat. Additionally, the algebra is wrong: from E[e_v(t+1)] = E[e_v(t)] - alpha E[RS_uw] RS*_v, the manuscript concludes E[e_v(t+1)] = (1 - alpha E[RS_uw] RS*_v) E[e(t)]. The second term is a constant, not proportional to E[e(t)], so the recurrence does not contract; it would drive the expected error to negative values, not to zero. Thus the main theoretical claim is unsupported.
  2. [Sections 3.4 and 4.4] The reputation feedback m_{j,k} = Omega(M(theta_j, theta_k)) - Omega(M(theta_j)) measures the change in peer j's local concordance when incorporating client k's update. With heterogeneous local data and censoring, an update can improve local concordance on peer j's data while degrading the global model's held-out concordance. No experiment in Section 5 measures the correlation between m_{j,k} and held-out global C-index; the reported reputation evolution plots (Figures 3-5) only show that scores change, not that they track contribution quality. Without such evidence, the central claim that the mechanism 'down-weighs noisy client updates' is not established.
  3. [Section 4.6] The DP accuracy guarantee assumes Omega is L_Omega-Lipschitz and that the global loss is L_k-smooth. For the concordance index, which is a step-function-based ranking statistic, Lipschitz continuity is nontrivial and is not argued or referenced; the Cox partial likelihood's smoothness in the reputation-weighted objective is also not verified. The bounds in Section 4.6.2 are therefore conditional on unverified assumptions. In addition, the 'Privacy and Accuracy Guarantee' in Section 4.4 refers to Section 4.6 for proof, but the proof only provides error bounds in terms of these assumptions and does not connect to an end-to-end privacy-utility trade-off.
  4. [Section 5.2.1, Table 2] The claim in the Abstract and Section 5.2.1 that the method 'consistently achieves high and stable C-index values... outperforming FL methods that lack a reputation system' is not supported by Table 2: the no-reputation baseline achieves 0.660216 in round 1, close to the proposed method's 0.661218, and TFFL reaches 0.6701 in round 7, higher than any value reported for the proposed method. The reported differences are not accompanied by confidence intervals or multiple-seed statistics, so it is unclear whether they are significant. The claim should be restated as 'on average' or supported with variance estimates.
minor comments (6)
  1. [Sections 3.3 and 5.1.2] The threat model defines noise added to feature vectors, but the experiments state noise is injected into model parameters or reputation messages; clarify which attack surface is used in each experiment.
  2. [Algorithm 1, line 8] The formula for selection probability is written with an ambiguous juxtaposition; it should be P(i) = RSi(r) divided by the sum of RS_j(r) over j in C_k.
  3. [Figure 3 caption] The caption says 'three plots' but lists four epsilon_max values; correct the count.
  4. [Section 4.4] The sentence 'ensures that clients do not gain knowledge of their peers' data or local model parameters' is too strong: clients receive DP-noised parameter vectors from peers, so they do gain information about noised parameters; suggest rephrasing to 'do not learn exact local parameters.'
  5. [Table 1] The table header 'Thonest' should be 'T_honest', and 'reputation Stability' should be 'reputation stability.'
  6. [Notation throughout] The symbol alpha is used both for the reputation learning rate in Eq. (4) and for the noise scaling in Section 3.3 (alpha_i); renaming one of the two would avoid confusion.

Circularity Check

1 steps flagged · score 6.0 of 10

Section 4.5's convergence proof assumes the very reliability it claims to establish: the peer-feedback signal is set equal to true reliability plus noise, so the claimed convergence is an input, not a derivation.

  1. self definitional [Section 4.5 (Convergence Analysis of Peer-Driven Reputation Estimation), Case 1 proof]
    "Given reliable feedback, the reputation estimated by a client converges to the true reliability of the peer ... Given the reputation update rule: RS uv(t+1) = RS uv(t) + α × RS uw(t) × m wv(t), where m wv(t) = RS ∗ v + η, and η is a zero-mean noise term with finite variance"

    The Lemma's conclusion is that reputation RS_uv(t) converges to the peer's true reliability RS*_v. The proof's substantive premise is that the feedback signal m_wv(t) equals RS*_v plus zero-mean noise. But m_wv is defined operationally in Sections 3.4 and 4.4 as the change in peer w's local C-index when v's update is included, i.e. m_wv = Ω(M(θ_w, θ_v)) − Ω(M(θ_w)). No argument shows this C-index delta is an unbiased estimator of RS*_v; the proof simply substitutes that identity. Thus the convergence is an input assumption, not a derived theorem about the reputation mechanism. The claimed contraction is also algebraically unsupported: E[e(t+1)] = E[e(t)] − αE[RS_uw]RS*_v is not equal to (1 − αE[RS_uw]RS*_v)E[e(t)], so the conclusion does not follow even from the assumed identity.

full rationale

The central theoretical support for the reputation mechanism is circular in Section 4.5: to prove that reputation scores converge to true reliability, the paper assumes that the feedback signal is true reliability plus zero-mean noise. The operational definition of that feedback, however, is a local C-index improvement, and no step in the paper connects local C-index improvements to an independent notion of global contribution quality. As a result, the convergence claim reduces to the assumption that the feedback is already reliable. The empirical comparisons in Tables 2 and 3 use a held-out global evaluation dataset against TFFL and no-reputation baselines, so those results are not circular, and the DP accuracy bounds in Section 4.6 are derived from Lipschitz and bounded-norm assumptions rather than from the conclusion. The self-citations to the authors' prior work are background and not load-bearing. Because one central theoretical claim is circular while the experimental and DP-privacy content retains independent substance, the overall circularity score is 6.

Assumptions & free parameters 5 free parameters · 6 assumptions · 0 invented entities

The method relies on standard FL, Cox, and DP machinery plus the paper's own reputation update. The key unproven inputs are the Lipschitz and smoothness assumptions, the faithfulness of C-index improvement as a contribution signal, and the circular feedback assumption in the convergence proof.

free parameters (5)
  • reputation learning rate alpha
    Controls the magnitude of reputation updates; sensitivity is tested in Section 5.1.4 but no selection rule or held-out tuning is reported.
  • clustering regularization lambda
    Balances Euclidean completeness clustering against intra-cluster C-index; tested at 0.05, 0.1, 0.2, and 0.5 without a selection criterion in Section 5.1.5.
  • DP noise budget (Q, epsilon, delta) and sigma
    The privacy-utility trade-off is assumed but numerical values are not reported, so the DP claims are not reproducible.
  • number of clusters c
    Algorithm 1 requires c clusters but provides no method for choosing c.
  • threat model parameters T_honest, T_ramp, epsilon_max
    These define the adversarial scenario in Section 3.3, but the values used in experiments are not fully listed, making Figures 3 to 5 hard to interpret.
assumptions (6)
  • domain assumption Cox proportional hazards assumption: covariates have multiplicative and time-independent effects on hazard
    Used throughout as the survival model and is not validated on SEER data.
  • standard math Gaussian mechanism provides (epsilon, delta)-differential privacy with the stated sigma
    Standard theorem, cited as [10,11]; applies only if sensitivity is bounded correctly, which the paper does not fully verify for Cox coefficients.
  • domain assumption C-index is an appropriate measure of model quality
    Used as the objective in both reputation updates and clustering; no external validation is given.
  • ad hoc to paper Omega is L_Omega-Lipschitz and the global loss is L_k-smooth
    Stated in Section 4.6.1 without verification and needed for the DP accuracy bounds.
  • ad hoc to paper Feedback m_wv(t) equals true reliability plus zero-mean noise
    Section 4.5 Case 1 assumes the convergence target inside the feedback, making the proof circular.
  • domain assumption Peer local C-index improvement is a faithful signal of global contribution quality
    Underlies the reputation update in Sections 3.4 and 4.4 but is never tested under data heterogeneity.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Enhancing Federated Survival Analysis through Peer-Driven Client Reputation in Healthcare." pith.science (2026). https://pith.science/paper/QA7TVAJY

@misc{pith2026250516190,
  author       = {Pith},
  title        = {Pith review of: Enhancing Federated Survival Analysis through Peer-Driven Client Reputation in Healthcare},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/QA7TVAJY}},
  note         = {Machine review of arXiv:2505.16190}
}
read the original abstract

Federated Learning (FL) holds great promise for digital health by enabling collaborative model training without compromising patient data privacy. However, heterogeneity across institutions, lack of sustained reputation, and unreliable contributions remain major challenges. In this paper, we propose a robust, peer-driven reputation mechanism for federated healthcare that employs a hybrid communication model to integrate decentralized peer feedback with clustering-based noise handling to enhance model aggregation. Crucially, our approach decouples the federated aggregation and reputation mechanisms by applying differential privacy to client-side model updates before sharing them for peer evaluation. This ensures sensitive information remains protected during reputation computation, while unaltered updates are sent to the server for global model training. Using the Cox Proportional Hazards model for survival analysis across multiple federated nodes, our framework addresses both data heterogeneity and reputation deficit by dynamically adjusting trust scores based on local performance improvements measured via the concordance index. Experimental evaluations on both synthetic datasets and the SEER dataset demonstrate that our method consistently achieves high and stable C-index values, effectively down-weighing noisy client updates and outperforming FL methods that lack a reputation system.

Figures

Figures reproduced from arXiv: 2505.16190 by the authors.

Figure 1
Figure 1. The proposed approach combining server-client and peer-to-peer data flow for model [PITH_FULL_IMAGE:figures/full_fig_p005_1.png] view at source ↗
Figure 2
Figure 2. The detailed workflow of the proposed reputation-enhanced federated survival analysis [PITH_FULL_IMAGE:figures/full_fig_p008_2.png] view at source ↗
Figure 3
Figure 3. Reputation score evolution over rounds for different noise levels ( [PITH_FULL_IMAGE:figures/full_fig_p014_3.png] view at source ↗
Figures from the paper (4 more)
Figure 4
Figure 4. Figure 4: Reputation score evolution over rounds for different values of neighbor importance during [PITH_FULL_IMAGE:figures/full_fig_p015_4.png]
Figure 5
Figure 5. Figure 5: Reputation score evolution over rounds for different values of the importance of intra [PITH_FULL_IMAGE:figures/full_fig_p016_5.png]
Figure 6
Figure 6. Figure 6: Impact of varying reputation update frequency ( [PITH_FULL_IMAGE:figures/full_fig_p017_6.png]
Figure 7
Figure 7. Figure 7: Average reputation score evolution with a static noise bias of 0 [PITH_FULL_IMAGE:figures/full_fig_p018_7.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

32 extracted references · 25 canonical work pages

  1. [1]

    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,” inArtificial intelligence and statistics. PMLR, 2017, pp. 1273–1282

    2017

  2. [2]

    Competition in healthcare: good, bad or ugly?

    M. Goddard, “Competition in healthcare: good, bad or ugly?”International journal of health policy and management, vol. 4, no. 9, p. 567, 2015

    work page 2015

  3. [3]

    Using geographic location-based public health features in survival analysis,

    N. Seidi, A. Tripathy, and S. K. Das, “Using geographic location-based public health features in survival analysis,” in2023 IEEE/ACM Conference on Connected Health: Applications, Systems and Engineering Technologies (CHASE), 2023, pp. 80–91

    work page 2023

  4. [4]

    Clusterfl: A clustering-based federated learning system for human activity recognition,

    X. Ouyang, Z. Xie, J. Zhou, G. Xing, and J. Huang, “Clusterfl: A clustering-based federated learning system for human activity recognition,”ACM Transactions on Sensor Networks, vol. 19, no. 1, pp. 1–32, 2022

    work page 2022

  5. [5]

    Dynamic clustering in federated learning,

    Y. Kim, E. Al Hakim, J. Haraldson, H. Eriksson, J. M. B. da Silva, and C. Fischione, “Dynamic clustering in federated learning,” inICC 2021-IEEE International Conference on Communi- cations. IEEE, 2021, pp. 1–6

    work page 2021

  6. [6]

    Addressing data heterogeneity in federated learning of cox proportional hazards models,

    N. Seidi, S. Roy, S. K. Das, and A. Tripathy, “Addressing data heterogeneity in federated learning of cox proportional hazards models,” in2024 IEEE International Conference on E- health Networking, Application & Services (HealthCom), 2024, pp. 1–7

    work page 2024

  7. [7]

    Incentive mechanism for reliable federated learning: A joint optimization approach to combining reputation and contract theory,

    J. Kang, Z. Xiong, D. Niyato, S. Xie, and J. Zhang, “Incentive mechanism for reliable federated learning: A joint optimization approach to combining reputation and contract theory,”IEEE Internet of Things Journal, vol. 6, no. 6, pp. 10 700–10 714, 2019

    work page 2019

  8. [8]

    A novel reputation-aware client selection scheme for federated learning within mobile environments,

    Y. Wang and B. Kantarci, “A novel reputation-aware client selection scheme for federated learning within mobile environments,” in2020 IEEE 25th International Workshop on Com- puter Aided Modeling and Design of Communication Links and Networks (CAMAD). IEEE, 2020, pp. 1–6

    work page 2020

Show all 32 references
  1. [9]

    Trustworthy and fair federated learning via reputation-based consensus and adaptive incentives,

    M. M. Rashid, Y. Xiang, M. P. Uddin, J. Tang, K. Sood, and L. Gao, “Trustworthy and fair federated learning via reputation-based consensus and adaptive incentives,”IEEE Transac- tions on Information Forensics and Security, 2025

    2025

  2. [10]

    Deep learning with differential privacy,

    M. Abadi, A. Chu, I. Goodfellow, H. B. McMahan, I. Mironov, K. Talwar, and L. Zhang, “Deep learning with differential privacy,” inProceedings of the 2016 ACM SIGSAC conference on computer and communications security, 2016, pp. 308–318

    2016

  3. [11]

    Calibrating noise to sensitivity in private data analysis,

    C. Dwork, F. McSherry, K. Nissim, and A. Smith, “Calibrating noise to sensitivity in private data analysis,” inTheory of Cryptography: Third Theory of Cryptography Conference, TCC 2006, New York, NY, USA, March 4-7, 2006. Proceedings 3. Springer, 2006, pp. 265–284. 20

    2006

  4. [12]

    Regression models and life-tables,

    D. R. Cox, “Regression models and life-tables,”Journal of the Royal Statistical Society: Series B (Methodological), vol. 34, no. 2, pp. 187–202, 1972

    1972

  5. [13]

    Trustworthy federated learning: Privacy, security, and beyond,

    C. Chen, J. Liu, and H. Tan, “Trustworthy federated learning: Privacy, security, and beyond,” Knowledge and Information Systems, 2024

    2024

  6. [14]

    Byzantine-robust decentralized federated learning,

    M. Fang, Z. Zhang, Hairi, P. Khanduri, J. Liu, S. Lu, Y. Liu, and N. Gong, “Byzantine-robust decentralized federated learning,”Proceedings of the ACM SIGSAC Conference on Computer and Communications Security, pp. 2874–2888, 2024

    2024

  7. [15]

    Defta: A plug-and-play peer-to-peer decentralized federated learning framework,

    Y. Zhou, M. Shi, Y. Tian, Q. Ye, and J. Lv, “Defta: A plug-and-play peer-to-peer decentralized federated learning framework,”Information Sciences, vol. 670, p. 120582, 2024

    2024

  8. [16]

    Machine learning with ad- versaries: Byzantine tolerant gradient descent,

    P. Blanchard, E. M. El Mhamdi, R. Guerraoui, and J. Stainer, “Machine learning with ad- versaries: Byzantine tolerant gradient descent,”Advances in neural information processing systems, vol. 30, 2017

    2017

  9. [17]

    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,” inInternational conference on machine learning. Pmlr, 2018, pp. 5650–5659

    2018

  10. [18]

    The hidden vulnerability of distributed learning in byzan- tium,

    R. Guerraoui, S. Rouaultet al., “The hidden vulnerability of distributed learning in byzan- tium,” inInternational conference on machine learning. PMLR, 2018, pp. 3521–3530

    2018

  11. [19]

    F. E. Harrellet al.,Regression modeling strategies: with applications to linear models, logistic regression, and survival analysis. Springer, 2001, vol. 608

    2001

  12. [20]

    Federated learning using peer-to-peer network for decentralized orchestration of model weights,

    M. R. Behera, S. Upadhyay, S. Shetty, and R. Otter, “Federated learning using peer-to-peer network for decentralized orchestration of model weights,”Authorea Preprints, 2023

    2023

  13. [21]

    scikit-survival: A library for time-to-event analysis built on top of scikit-learn,

    S. P¨ olster, “scikit-survival: A library for time-to-event analysis built on top of scikit-learn,” J. of Machine Learning Research, vol. 21, no. 212, pp. 1–6, 2020. [Online]. Available: http://jmlr.org/papers/v21/20-729.html

    2020

  14. [22]

    Analysis of survival data under the proportional hazards model,

    N. E. Breslow, “Analysis of survival data under the proportional hazards model,”International Statistical Review/Revue Internationale de Statistique, pp. 45–57, 1975

    1975

  15. [23]

    Learning differentially private recur- rent language models,

    H. B. McMahan, D. Ramage, K. Talwar, and L. Zhang, “Learning differentially private recur- rent language models,”arXiv preprint arXiv:1710.06963, 2017

    2017 arXiv

  16. [24]

    Differentially private federated learning: A client level perspective,

    R. C. Geyer, T. Klein, and M. Nabi, “Differentially private federated learning: A client level perspective,”arXiv preprint arXiv:1712.07557, 2017

    2017 arXiv

  17. [25]

    R. J. Little and D. B. Rubin,Statistical analysis with missing data. John Wiley & Sons, 2019, vol. 793

    2019

  18. [26]

    Federated survival analysis with discrete-time cox models,

    M. Andreux, A. Manoel, R. Menuet, C. Saillard, and C. Simpson, “Federated survival analysis with discrete-time cox models,”arXiv preprint arXiv:2006.08997, 2020

    2006 arXiv

  19. [27]

    Explaining and harnessing adversarial examples,

    I. J. Goodfellow, J. Shlens, and C. Szegedy, “Explaining and harnessing adversarial examples,” inInternational Conference on Learning Representations (ICLR), 2014. 21

    2014

  20. [28]

    Delving into transferable adversarial examples and black-box attacks,

    Y. Liu, X. Chen, C. Liu, and D. Song, “Delving into transferable adversarial examples and black-box attacks,” inInternational Conference on Learning Representations, 2017. [Online]. Available: https://openreview.net/forum?id=Sys6GJqxl

    2017

  21. [29]

    Practical poissonian-gaussian noise modeling and fitting for single-image raw-data,

    A. Foi, M. Trimeche, V. Katkovnik, and K. Egiazarian, “Practical poissonian-gaussian noise modeling and fitting for single-image raw-data,”IEEE transactions on image processing, vol. 17, no. 10, pp. 1737–1754, 2008

    2008

  22. [30]

    R. C. Gonzalez and R. E. Woods,Digital Image Processing. Prentice Hall, 2008

    2008

  23. [31]

    P. J. Huber,Robust Statistics. Wiley, 2011

    2011

  24. [32]

    SEER Research Data 1975–2021 when using SEER*Stat: Surveillance, Epidemiology, and End Results (SEER) Program,

    S. R. P. National Cancer Institute, DCCPS, “SEER Research Data 1975–2021 when using SEER*Stat: Surveillance, Epidemiology, and End Results (SEER) Program,” SEER*Stat Database: Incidence - SEER Research Data, 8 Registries, Nov 2023 Sub (1975–2021) - Linked To County Attributes ...

    1975

Pith tools

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