Pith. sign in

REVIEW 2 major objections 39 references

Unlearning to Protect: A Distilled Reinforcement Learning Framework with Privacy-Preserving Feature Unlearning and XAI for IoT Security

T0 review · 2 major / 0 minor · reviewed 2026-07-10 · grok-4.5

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

Pith's one-line read A distilled A2C student detects IoT botnet traffic at 99.6% accuracy with 2370 FLOPS and reversible post-hoc feature unlearning.

desk verdict Practical packaging of A2C distillation plus crude weight-zeroing unlearning for edge botnet detection, with strong empirical numbers and code, but the GDPR claim rests on an incomplete unlearning argument. read the letter →

arxiv 2607.07635 v2 pith:DDKMLCAS submitted 2026-07-08 cs.CR

Md. Nahid Hasan , Md. Golam Rabiul Alam This is my paper · ORCID
classification cs.CR
keywords botnetdetectionIoTsecurityknowledgedistillationfeatureunlearningA2CLIMEedgedeploymentGDPR
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

IoT devices remain exposed to botnets because most AI detectors are too heavy for edge hardware and cannot forget sensitive or outdated features without a full retrain. DiRLU answers both problems at once: a large A2C teacher is distilled into a compact A2C student that still reaches 99.60% accuracy and 99.80% F1 on a 25% slice of BoT-IoT, while a one-line weight edit (zeroing the first-layer column of a chosen feature) removes that feature’s influence. Restoring the same weights recovers the original scores, proving the edit is reversible. LIME explanations confirm the forgotten feature no longer appears among the top drivers, and the student needs only 2370 FLOPS—about 3.87 times fewer than the cited KronNet baseline—so the whole pipeline fits resource-constrained gateways and cameras.

What carries the argument

Post-hoc weight modification (PHWM): after training, set the entire first-layer weight column W:,j = 0 for the target feature so that feature never again reaches any later layer; the same column can later be restored from a saved copy.

What would settle it

After PHWM is applied to a high-influence feature such as flgs, extract LIME explanations (or gradient attributions) on held-out attack samples; if that feature still ranks among the top contributors or if accuracy does not drop then recover upon weight restoration, the unlearning claim fails.

Watch

Extended reading notes

Core claim

Knowledge distillation from an attention-augmented A2C teacher into a smaller A2C student yields a 2370-FLOP detector that matches the teacher’s 99.6% accuracy on BoT-IoT, while a post-hoc zeroing of the corresponding first-layer weight column cleanly erases any chosen feature’s contribution; reloading the original weights fully restores performance, satisfying a practical “right-to-be-forgotten” requirement without retraining.

Load-bearing premise

Zeroing only the first-layer weight column for a feature is enough to erase every trace of that feature from the model’s decisions and from any residual information that might remain in later layers, attention maps, or the distillation process itself.

Editorial extensions

If this is right

  • Edge IoT gateways can run a 3 k-parameter botnet detector at full accuracy without cloud offload.
  • Operators can comply with GDPR-style erasure requests by a single matrix write rather than a multi-hour retrain.
  • Because the unlearning step is reversible, temporary privacy redactions can be rolled back when the legal basis for processing reappears.
  • The same teacher–student + PHWM pattern can be reused for any tabular network-traffic classifier that needs both size reduction and selective forgetting.

Reading between the lines

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

  • If first-layer zeroing is later shown to leave residual traces in attention or critic heads, a stronger unlearning operator (e.g., also zeroing corresponding attention weights) would be required before the GDPR claim can be certified.
  • The same lightweight student could serve as the local model in a federated IoT setting, allowing devices to unlearn a feature without ever shipping raw traffic off-device.
  • Because the student already uses an attention mechanism, replacing LIME with attention-rollout maps could give real-time, zero-extra-cost explanations on the device itself.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

2 major / 0 minor

Summary. The paper proposes DiRLU, a knowledge-distillation framework in which both teacher and student are A2C actor–critic networks with feature-level attention, trained on 25% (and 30%) of the BoT-IoT dataset after SMOTE balancing. The student reaches 99.60% accuracy / 99.80% F1 while using only 2,370 FLOPS. A post-hoc weight-modification (PHWM) step zeros the first dense-layer column corresponding to a chosen feature (flgs), after which accuracy drops only modestly and is fully restored when the original weights are re-inserted. LIME explanations are supplied for a single attack instance before and after unlearning. The authors claim the combination yields a lightweight, GDPR-aligned (“right to be forgotten”) IoT botnet detector.

Significance. If the efficiency numbers hold under realistic edge constraints and the unlearning claim can be made rigorous, the work would supply a practical, reversible privacy mechanism for resource-constrained IoT security models—something largely missing from prior botnet detectors that rely on full retraining. The public code link, the use of a larger data slice than many BoT-IoT baselines, and the explicit FLOPS comparison with KronNet are concrete strengths that make the efficiency claim falsifiable and useful to the community.

major comments (2)
  1. Section III-H (Eqs. 12–14) and Algorithm 1 assert that setting the first dense-layer column W:,j = 0 completely removes feature xj because later layers only see z. Both teacher and student, however, employ a feature-level attention mechanism whose weighted inputs are concatenated with the original features before the shared dense layers (Sections III-F/G). Attention weights, residual paths through the critic, and soft-target distillation can therefore still encode information about the “forgotten” feature. The only evidence offered is a modest accuracy/F1 drop and a single LIME instance in which flgs disappears from the top ranks. No mutual-information, membership-inference, or reconstruction test is reported. Consequently the GDPR “right to be forgotten” claim remains informal and load-bearing for the paper’s privacy contribution.
  2. The evaluation relies on a single 70/20/10 split of a heavily SMOTE-augmented 25% (and 30%) slice; no standard deviations across random seeds, no cross-validation, and no statistical significance tests accompany the 99.6% accuracy figures (Table VII). Given the extreme original imbalance (≈98:2) and the free parameters (T=2, α=0.5, decision threshold 0.8, SMOTE ratio), the reported margins over baselines cannot be assessed for robustness. At least multi-seed means and confidence intervals are required before the efficiency–accuracy trade-off can be trusted.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: empirical measurements of accuracy/F1/FLOPS and post-hoc weight zeroing on a public dataset; no claimed prediction reduces to a fitted free parameter or self-definition by construction.

full rationale

The paper presents an engineering framework (A2C teacher-student distillation + first-layer weight zeroing for unlearning + LIME) evaluated empirically on a 25% BoT-IoT subset. Accuracy (99.60%), F1 (99.80%), FLOPS (2370), and the modest drop/restore after zeroing W:,j are measured outcomes on held-out data, not derived quantities forced by construction from fitted inputs. Equations 12-14 simply define the unlearning operator (set column to zero so z ignores xj); the subsequent accuracy/LIME checks are independent empirical tests of its effect, not tautological restatements. Hyper-parameters (T=2, α=0.5, threshold 0.8) are chosen by hand but do not algebraically produce the reported metrics. No load-bearing uniqueness theorem, self-citation chain, or renamed known result appears. Reversibility after weight restoration is expected by construction yet is presented only as a secondary sanity check, not as a central prediction. The work is therefore self-contained against external benchmarks with no circular reduction.

Assumptions & free parameters 5 free parameters · 4 assumptions · 2 invented entities

The central empirical claims rest on standard RL and distillation mathematics, the modeling choice that first-layer weight zeroing constitutes unlearning, several hand-chosen hyper-parameters, and the synthetic balancing of an extremely imbalanced public dataset. No new physical entities are postulated.

free parameters (5)
  • distillation temperature T = 2.0
    Set to 2.0 by hand to soften teacher logits; directly scales the KL term in the student loss.
  • loss mixing weight alpha = 0.5
    Balances cross-entropy versus distillation KL; fixed at 0.5 without ablation reported as optimal.
  • actor decision threshold = 0.8
    Probability >0.8 is treated as attack; chosen without ROC analysis shown.
  • SMOTE oversampling ratio = 67:33 after SMOTE
    Transforms original ~98:2 attack:benign ratio into 67:33; the exact synthetic count is a free modeling choice that affects all metrics.
  • attention entropy penalty lambda
    Regularization strength on attention weights; value not numerically specified beyond presence in loss.
assumptions (4)
  • domain assumption Actor-critic (A2C) policy and value updates with advantage estimates converge to a useful classifier for network-flow features.
    Standard RL theory is assumed to transfer to the supervised-like botnet classification setting used throughout Sections III-F and III-G.
  • ad hoc to paper Setting the first-layer weight column corresponding to feature j exactly to zero removes all influence of that feature from every subsequent layer and from the final decision.
    Core of the post-hoc unlearning claim (Eqs. 12-14 and Algorithm 1); not derived from a formal unlearning guarantee.
  • domain assumption SMOTE-generated synthetic minority samples preserve the true decision boundary of real benign traffic.
    Required for the reported high recall on the minority class after balancing (Section III-D).
  • domain assumption A randomly chosen 25% (or 30%) subset of BoT-IoT is distributionally representative of the full 72 M-record corpus and of real IoT traffic.
    Justifies all accuracy claims relative to prior work that used 5% slices (Section III-C).
invented entities (2)
  • DiRLU framework
    purpose: Name for the end-to-end pipeline of A2C distillation + PHWM unlearning + LIME.
    Organizational label; no independent physical existence claimed.
  • Post-hoc weight modification (PHWM) unlearning
    purpose: Mechanism that zeros selected first-layer weights to forget a feature without retraining.
    Presented as the privacy-preserving contribution; efficacy is measured only inside the same models.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Unlearning to Protect: A Distilled Reinforcement Learning Framework with Privacy-Preserving Feature Unlearning and XAI for IoT Security." pith.science (2026). https://pith.science/paper/DDKMLCAS

@misc{pith2026260707635,
  author       = {Pith},
  title        = {Pith review of: Unlearning to Protect: A Distilled Reinforcement Learning Framework with Privacy-Preserving Feature Unlearning and XAI for IoT Security},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/DDKMLCAS}},
  note         = {Machine review of arXiv:2607.07635}
}
read the original abstract

Botnets pose a significant cybersecurity threat, enabling attacks such as DDoS, data theft, and service disruptions on IoT devices. These devices often lack built-in botnet traffic filtering, leaving them highly exposed. Existing AI-based solutions improve detection capabilities but have limitations: (i) they are too heavy for IoT deployment, and (ii) they lack unlearning capabilities to forget sensitive or outdated features without retraining. To address these challenges, we propose DiRLU, a lightweight, reinforcement learning driven framework, while ensuring privacy by selectively unlearning sensitive or outdated features without requiring retraining. The framework leverages knowledge distillation to transfer knowledge from a teacher model into a lightweight student model, with both models trained using A2C. A post-hoc unlearning mechanism modifies weights to remove targeted features, while restored features show negligible performance loss, confirming reversibility. Unlike many benchmark models that used only 5% of the BoT-IoT dataset, this research leverages 25%, allowing us to develop a strong teacher model. Both the teacher and student models were trained using the A2C reinforcement learning algorithm, achieving impressive results, with the student model achieving 99.60% accuracy and a 99.80% F1 score. To enhance transparency, we integrated Explainable AI (XAI), particularly LIME, which helps interpret the model's decisions and identify the key features influencing its predictions. Moreover, DiRLU requires only 2,370 FLOPS, approximately 3.87x more efficient than the state-of-the-art model, highlighting its efficiency for edge deployment. DiRLU combines efficiency with privacy, aligning with GDPR standards (right to be forgotten) to provide practical and scalable IoT security solution.

Figures

Figures reproduced from arXiv: 2607.07635 by the authors.

Figure 1
Figure 1. A high-level overview of the proposed research [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. A detailed architecture of the DiRLU framework, where the teacher transfers knowledge to the student via knowledge [PITH_FULL_IMAGE:figures/full_fig_p006_2.png] view at source ↗
Figure 3
Figure 3. Schematic overview of the proposed DiRLU framework based on knowledge distillation and post-hoc weight modification [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figures from the paper (16 more)
Figure 4
Figure 4. Figure 4: Comparison of before and after applying SMOTE [PITH_FULL_IMAGE:figures/full_fig_p007_4.png]
Figure 5
Figure 5. Figure 5: Actor-Critic model architecture From equations 2-7, we have described the mathematical formulas of the actor-critic model, including both forward propagation and backpropagation update rules. πθ(a|s) = P(a|s; θ) (2) [PITH_FULL_IMAGE:figures/full_fig_p008_5.png]
Figure 6
Figure 6. Figure 6: Architecture of knowledge distillation (KD) framework [PITH_FULL_IMAGE:figures/full_fig_p009_6.png]
Figure 7
Figure 7. Figure 7: Architecture of feature unlearning process [PITH_FULL_IMAGE:figures/full_fig_p011_7.png]
Figure 8
Figure 8. Figure 8: Training accuracy and loss at each epoch of Teacher [PITH_FULL_IMAGE:figures/full_fig_p013_8.png]
Figure 9
Figure 9. Figure 9: Confusion matrix of Teacher model illustrates the convergence trends of training and validation accuracy and loss across epochs. In the final evaluation, the student model achieved an impressive accuracy of 99.55%, an F1 score of 99.78%, and a minimal test loss of 0.04…
Figure 10
Figure 10. Figure 10: Training accuracy and loss at each epoch of Student [PITH_FULL_IMAGE:figures/full_fig_p014_10.png]
Figure 13
Figure 13. Figure 13: Confusion matrix of Teacher unlearn model [PITH_FULL_IMAGE:figures/full_fig_p015_13.png]
Figure 14
Figure 14. Figure 14: Comparison of Student and Student unlearn models: [PITH_FULL_IMAGE:figures/full_fig_p015_14.png]
Figure 15
Figure 15. Figure 15: Confusion matrix of Student unlearn model [PITH_FULL_IMAGE:figures/full_fig_p015_15.png]
Figure 17
Figure 17. Figure 17: XAI interpretation of the Teacher model’s prediction [PITH_FULL_IMAGE:figures/full_fig_p016_17.png]
Figure 18
Figure 18. Figure 18: XAI interpretation of top 12 features impacts for the [PITH_FULL_IMAGE:figures/full_fig_p016_18.png]
Figure 19
Figure 19. Figure 19: XAI interpretation of the Student model’s prediction [PITH_FULL_IMAGE:figures/full_fig_p016_19.png]
Figure 22
Figure 22. Figure 22: XAI interpretation of top 12 features impacts for the [PITH_FULL_IMAGE:figures/full_fig_p016_22.png]
Figure 23
Figure 23. Figure 23: XAI interpretation of the Student Unlearn model’s [PITH_FULL_IMAGE:figures/full_fig_p017_23.png]
Figure 24
Figure 24. Figure 24: XAI interpretation of top 12 features impacts for the [PITH_FULL_IMAGE:figures/full_fig_p017_24.png]

Review history (2 revisions) →

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

39 extracted references · 39 canonical work pages

  1. [1]

    What is a botnet?

    P. A. Networks, “What is a botnet?” 2024, last checked on 12.06.2024. [Online]. Available: https://www.paloaltonetworks.com/ cyberpedia/what-is-botnet

    work page 2024

  2. [2]

    Mirai ddos attack on dyn case study,

    CoverLink, “Mirai ddos attack on dyn case study,” 2023, accessed: 2024-11-06. [Online]. Available: https://coverlink.com/case-study/ mirai-ddos-attack-on-dyn/

    work page 2023

  3. [3]

    What is gdpr?

    GDPR.eu, “What is gdpr?” 2024, last checked on 19.07.2025. [Online]. Available: https://gdpr.eu/what-is-gdpr/

    work page 2024

  4. [4]

    Detecting information theft attacks in the bot-iot dataset,

    J. L. Leevy, J. Hancock, T. M. Khoshgoftaar, and J. Peterson, “Detecting information theft attacks in the bot-iot dataset,” in2021 20th IEEE Inter- national Conference on Machine Learning and Applications (ICMLA). IEEE, 2021, pp. 807–812

    work page 2021

  5. [6]

    An intrusion detection system using bot-iot,

    S. Alosaimi and S. M. Almutairi, “An intrusion detection system using bot-iot,”Applied Sciences, vol. 13, no. 9, p. 5427, 2023

    work page 2023

  6. [7]

    A novel method for malware detection using audio signal processing techniques,

    M. Farrokhmanesh and A. Hamzeh, “A novel method for malware detection using audio signal processing techniques,” in2016 Artificial Intelligence and Robotics (IRANOPEN). IEEE, 2016, pp. 85–91

    work page 2016

  7. [8]

    A survey of malware detection using deep learning,

    A. Bensaoud, J. Kalita, and M. Bensaoud, “A survey of malware detection using deep learning,”Machine Learning With Applications, vol. 16, p. 100546, 2024

    work page 2024

  8. [9]

    Machine Unlearning: A Comprehensive Survey

    W. Wang, Z. Tian, C. Zhang, and S. Yu, “Machine unlearning: A comprehensive survey,”arXiv preprint arXiv:2405.07406, 2024

    work page Pith review arXiv 2024

Show all 39 references
  1. [10]

    A lightweight malware detection model based on knowledge distillation

    C. Miao, L. Kou, J. Zhang, and G. Dong, “A lightweight malware detection model based on knowledge distillation.”Mathematics (2227- 7390), vol. 12, no. 24, 2024

    2024

  2. [11]

    Explainability in ai-based behavioral malware detection systems,

    A. Galli, V . La Gatta, V . Moscato, M. Postiglione, and G. Sperl `ı, “Explainability in ai-based behavioral malware detection systems,”Com- puters & Security, vol. 141, p. 103842, 2024

    2024

  3. [12]

    Machine learning-based early detection of iot botnets using network-edge traffic,

    A. Kumar, M. Shridhar, S. Swaminathan, and T. J. Lim, “Machine learning-based early detection of iot botnets using network-edge traffic,” Computers & Security, vol. 117, p. 102693, 2022

    2022

  4. [13]

    Paired: An explainable lightweight android malware detection system,

    M. M. Alani and A. I. Awad, “Paired: An explainable lightweight android malware detection system,”IEEE Access, vol. 10, pp. 73 214– 73 228, 2022

    2022

  5. [14]

    Assessing llms in malicious code deobfuscation of real-world malware campaigns,

    C. Patsakis, F. Casino, and N. Lykousas, “Assessing llms in malicious code deobfuscation of real-world malware campaigns,”Expert Systems with Applications, vol. 256, p. 124912, 2024

    2024

  6. [15]

    A privacy-enhanced framework with deep learning for botnet detection,

    G. Wu and X. Wang, “A privacy-enhanced framework with deep learning for botnet detection,”Cybersecurity, vol. 8, no. 1, p. 9, 2025

    2025

  7. [16]

    Reinforcement learning for iot security: A comprehensive survey,

    A. Uprety and D. B. Rawat, “Reinforcement learning for iot security: A comprehensive survey,”IEEE Internet of Things Journal, vol. 8, no. 11, pp. 8693–8706, 2021

    2021

  8. [17]

    Asynchronous advantage actor-critic (a3c) learning for cognitive network security,

    E. Muhati and D. B. Rawat, “Asynchronous advantage actor-critic (a3c) learning for cognitive network security,” in2021 Third IEEE International Conference on Trust, Privacy and Security in Intelligent Systems and Applications (TPS-ISA), 2021, pp. 106–113

    2021

  9. [18]

    Towards the development of realistic botnet dataset in the internet of things for network forensic analytics: Bot-iot dataset,

    N. Koroniotis, N. Moustafa, E. Sitnikova, and B. Turnbull, “Towards the development of realistic botnet dataset in the internet of things for network forensic analytics: Bot-iot dataset,”Future Generation Computer Systems, vol. 100, pp. 779–796, 2019

    2019

  10. [19]

    A comparative review of smote and adasyn in imbalanced data classification,

    J. Brandt and E. Lanz ´en, “A comparative review of smote and adasyn in imbalanced data classification,” 2021

    2021

  11. [20]

    Smote: synthetic minority over-sampling technique,

    N. V . Chawla, K. W. Bowyer, L. O. Hall, and W. P. Kegelmeyer, “Smote: synthetic minority over-sampling technique,”Journal of artificial intel- ligence research, vol. 16, pp. 321–357, 2002

    2002

  12. [21]

    Epileptic seizure detection using deep learning through min max scaler normalization,

    B. Deepa and K. Ramesh, “Epileptic seizure detection using deep learning through min max scaler normalization,”Int. J. Health Sci, vol. 6, pp. 10 981–10 996, 2022

    2022

  13. [22]

    Reinforcement learning,

    IBM, “Reinforcement learning,” 2023, accessed: 2024-11-06. [Online]. Available: https://www.ibm.com/think/topics/reinforcement-learning 19

    2023

  14. [23]

    Reinforcement learning for iot security: A comprehensive survey,

    A. Uprety and D. B. Rawat, “Reinforcement learning for iot security: A comprehensive survey,”IEEE Internet of Things Journal, vol. 8, no. 11, pp. 8693–8706, 2020

    2020

  15. [24]

    Actor-critic methods — tensorflow reinforcement learning,

    TensorFlow, “Actor-critic methods — tensorflow reinforcement learning,” 2023, accessed: 2024-11-06. [Online]. Available: https://www.tensorflow.org/tutorials/reinforcement learning/actor critic

    2023

  16. [25]

    Knowledge distillation: A survey,

    J. Gou, B. Yu, S. J. Maybank, and D. Tao, “Knowledge distillation: A survey,”International Journal of Computer Vision, vol. 129, no. 6, pp. 1789–1819, 2021

    2021

  17. [26]

    Improved knowledge distillation with teacher-student adaptation for efficient neural network compression,

    X. Chen, Z. Zhang, and L. Liu, “Improved knowledge distillation with teacher-student adaptation for efficient neural network compression,” IEEE Transactions on Neural Networks and Learning Systems, vol. 33, no. 9, pp. 4240–4251, 2022

    2022

  18. [27]

    The price of forgetting: Incentive mech- anism design for machine unlearning,

    Y . Cui and M. H. Cheung, “The price of forgetting: Incentive mech- anism design for machine unlearning,”IEEE Transactions on Mobile Computing, pp. 1–14, 2025

    2025

  19. [28]

    Learn to forget: Machine unlearning via neuron masking,

    Z. Ma, Y . Liu, X. Liu, J. Liu, J. Ma, and K. Ren, “Learn to forget: Machine unlearning via neuron masking,”IEEE Transactions on De- pendable and Secure Computing, vol. 20, no. 4, pp. 3194–3207, 2023

    2023

  20. [29]

    Machine unlearning by reversing the continual learning,

    Y . Zhang, Z. Lu, F. Zhang, H. Wang, and S. Li, “Machine unlearning by reversing the continual learning,”Applied Sciences, vol. 13, no. 16, p. 9341, 2023

    2023

  21. [30]

    Post-hoc spurious correlation neutralization with single-weight fictitious class unlearning,

    S. Hakemi, N. Akhtar, G. M. Hassan, and A. Mian, “Post-hoc spurious correlation neutralization with single-weight fictitious class unlearning,” arXiv preprint arXiv:2501.14182, 2025

    2025

  22. [31]

    ”why should i trust you?

    M. T. Ribeiro, S. Singh, and C. Guestrin, “”why should i trust you?” explaining the predictions of any classifier,” inProceedings of the 22nd ACM SIGKDD international conference on knowledge discovery and data mining, 2016, pp. 1135–1144

    2016

  23. [32]

    Anomaly detection and localization in nfv systems by utilizing masked- autoencoder and xai,

    S. S. Johari, N. Shahriar, M. Tornatore, R. Boutaba, and A. Saleh, “Anomaly detection and localization in nfv systems by utilizing masked- autoencoder and xai,”IEEE Transactions on Mobile Computing, pp. 1– 18, 2025

    2025

  24. [33]

    Explaining explanations: An overview of interpretability of machine learning,

    L. H. Gilpin, D. Bau, B. Z. Yuan, A. Bajwa, M. Specter, and L. Kagal, “Explaining explanations: An overview of interpretability of machine learning,” in2018 IEEE 5th International Conference on data science and advanced analytics (DSAA). IEEE, 2018, pp. 80–89

    2018

  25. [34]

    Kronnet a lightweight kronecker enhanced feed forward neural network for efficient iot intrusion detection,

    S. Ullah, J. Wu, M. M. Kamal, and A. K. J. Saudagar, “Kronnet a lightweight kronecker enhanced feed forward neural network for efficient iot intrusion detection,”Scientific Reports, vol. 15, no. 1, p. 20850, 2025

    2025

  26. [35]

    A lightweight intrusion detection method for iot based on deep learning and dynamic quantization. peerj comput. sci. 9, e1569 (2023)

    Z. Wang, H. Chen, S. Yang, X. Luo, D. Li, and J. Wang, “A lightweight intrusion detection method for iot based on deep learning and dynamic quantization. peerj comput. sci. 9, e1569 (2023).”

    2023

  27. [36]

    A lightweight network intrusion detec- tion method based on protocol-aware dynamic inverted residuals and a sliding-window multi-batch self-knowledge distillation strategy,

    S. Feng, S. Ma, and M. Ma, “A lightweight network intrusion detec- tion method based on protocol-aware dynamic inverted residuals and a sliding-window multi-batch self-knowledge distillation strategy,” in Proceedings of the 2nd Guangdong-Hong Kong-Macao Greater Bay Area Educat...

    2025

  28. [37]

    Lightweight intrusion detection model of the internet of things with hybrid cloud-fog computing,

    G. Zhao, Y . Wang, and J. Wang, “Lightweight intrusion detection model of the internet of things with hybrid cloud-fog computing,”Security and Communication Networks, vol. 2023, no. 1, p. 7107663, 2023

    2023

  29. [38]

    A lightweight in- trusion detection system with dynamic feature fusion federated learning for vehicular network security,

    J. Li, Y . Ma, J. Bai, C. Chen, T. Xu, and C. Ding, “A lightweight in- trusion detection system with dynamic feature fusion federated learning for vehicular network security,”Sensors, vol. 25, no. 15, p. 4622, 2025

    2025

  30. [39]

    Cer-feaun: Certified feature unlearning in vertical federated learning,

    Y . Wang, Z. Lu, Z. Liu, T. Li, Z. Chen, and W. Susilo, “Cer-feaun: Certified feature unlearning in vertical federated learning,”IEEE Trans- actions on Mobile Computing, pp. 1–13, 2025

    2025

  31. [40]

    Verifi: Towards verifiable federated unlearning,

    X. Gao, X. Ma, J. Wang, Y . Sun, B. Li, S. Ji, P. Cheng, and J. Chen, “Verifi: Towards verifiable federated unlearning,”IEEE Transactions on Dependable and Secure Computing, vol. 21, no. 6, pp. 5720–5736, 2024

    2024

Pith tools

Reviewed July 10, 2026 · model on record in the stance chip above.