Machine Unlearning for the XGBoost Model with Network Intrusion Datasets

Diana Magalh\~aes; Eva Maia; Isabel Pra\c{c}a; Jo\~ao Vitorino

arxiv: 2606.19220 · v1 · pith:FYUYNEQKnew · submitted 2026-06-17 · 💻 cs.LG · cs.AI

Machine Unlearning for the XGBoost Model with Network Intrusion Datasets

Diana Magalh\~aes , Eva Maia , Jo\~ao Vitorino , Isabel Pra\c{c}a This is my paper

Pith reviewed 2026-06-26 21:14 UTC · model grok-4.3

classification 💻 cs.LG cs.AI

keywords machine unlearningXGBoostnetwork intrusion detectiontabular dataIoT-23GeNISforgetting qualitydata removal

0 comments

The pith

XGBoost-Forget removes targeted data from network intrusion models without full retraining while keeping similar accuracy.

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

The paper introduces XGBoost-Forget as a method to delete specific data points from an XGBoost model trained on network intrusion detection tasks. It applies the method to two tabular datasets, IoT-23 and GeNIS, and measures how well the resulting model still detects intrusions. The approach runs much faster than retraining the model from scratch on the remaining data. A sympathetic reader would care because intrusion detection systems handle sensitive records that sometimes must be removed for privacy reasons, and repeated full retraining is expensive on large tabular sets. The work shows that the unlearned models stay close in performance to the originals across the tested metrics.

Core claim

XGBoost-Forget is an unlearning technique for the XGBoost model that removes the influence of chosen data points on tabular network intrusion datasets. When evaluated on IoT-23 and GeNIS, the method produces models whose predictive performance remains close to that of the original trained model while completing the unlearning step in significantly less time than would be required for full retraining from scratch. The evaluation uses separate measures for overall model performance, the speed of the unlearning operation, and the quality of the forgetting achieved.

What carries the argument

XGBoost-Forget, the unlearning procedure applied to the XGBoost gradient boosting model that selectively adjusts for removal of targeted training records.

If this is right

The same unlearning procedure can be applied to other tabular network intrusion datasets without changing the core approach.
Operators gain a practical way to honor data removal requests while avoiding the full cost of retraining an XGBoost intrusion detector.
Detection accuracy on unseen traffic stays nearly unchanged, so the operational value of the model is preserved after unlearning.
The method scales to the size of typical IoT and network traffic logs where full retraining would be prohibitive.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

Similar unlearning steps could be adapted for other tree ensemble models used in security monitoring beyond the two datasets tested here.
Integration into live intrusion detection pipelines might allow periodic removal of outdated or compromised records with low overhead.
The approach could lower the barrier for deploying gradient boosting models in regulated environments that require data erasure capabilities.

Load-bearing premise

The chosen metrics for performance, speed, and forgetting quality are enough to confirm that the influence of the removed data points has truly been eliminated rather than merely masked on the test set.

What would settle it

A verification step in which an auditor checks whether the unlearned model still produces outputs that encode information unique to the removed data points would directly test whether forgetting succeeded.

read the original abstract

Machine Unlearning (MU) has emerged as an important technique for removing specific data points from trained models without requiring full retraining. However, most existing MU research focuses on deep learning and image data, leaving a gap in the domain of network intrusion detection, which relies heavily on tabular data. This work introduces XGBoost-Forget, an unlearning approach for the XGBoost model, to address this gap. The approach is evaluated on two tabular Network Intrusion (NI) datasets, IoT-23 and GeNIS, using multiple metrics to assess model performance, unlearning efficiency, and forgetting quality. The results show that XGBoost-Forget maintains predictive performance close to the original model while providing significantly faster unlearning, demonstrating its potential for MU in tabular NI settings.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

The paper applies unlearning to XGBoost on two intrusion datasets and shows efficiency gains, but test-set performance alone does not confirm the data was actually forgotten.

read the letter

The main point is that this work takes standard machine unlearning techniques and runs them on XGBoost with the IoT-23 and GeNIS tabular datasets. It reports that the method is faster than retraining from scratch while accuracy and F1 stay close on held-out test data. That combination is new enough for the narrow area of tree models in network security.

The experiments are straightforward and use real datasets that matter for intrusion detection. The efficiency numbers are the clearest part of the contribution, and the authors keep the focus on practical metrics rather than new theory. For someone who needs a quick unlearning option for XGBoost in this domain, the results give a usable starting point.

The soft spot is the forgetting verification. Performance staying similar on the test set is compatible with the model still carrying information from the removed points through unchanged splits or leaf statistics. The abstract mentions forgetting quality metrics but supplies no numbers, baselines, or comparison to a model retrained only on the retain set. Without that direct check or a membership-inference test, the claim that the influence has been removed stays unproven. This is the load-bearing gap.

The paper is aimed at applied people working on ML for security who already use tree ensembles and need to handle deletion requests. It is not broad enough to interest the wider unlearning community. The thinking is clear and the setup is honest, so it deserves a serious referee who can push for the missing verification step. I would send it to review rather than desk reject.

Referee Report

2 major / 2 minor

Summary. The paper introduces XGBoost-Forget, an unlearning method for XGBoost models on tabular network intrusion datasets (IoT-23 and GeNIS). It evaluates the approach on metrics for model performance, unlearning efficiency, and forgetting quality, claiming that the method maintains predictive performance close to the original model while achieving significantly faster unlearning than full retraining.

Significance. If the central claim of verifiable forgetting holds, the work would address a relevant gap by extending machine unlearning to tree-based models on tabular security data. The empirical evaluation on two real-world network intrusion datasets provides practical grounding. However, the absence of quantitative results, baselines, and verification steps in the abstract reduces the assessed contribution at present.

major comments (2)

[Abstract] Abstract: The abstract reports positive results on two datasets but provides no quantitative numbers, error bars, baseline comparisons, or details on how forgetting quality was measured. This makes the central empirical claims impossible to assess from the given information.
[Evaluation] Evaluation section: The reported metrics (accuracy/F1 similarity to the original model and faster runtime) are compatible with the model retaining influence from the forget set via unchanged splits or leaf statistics. No comparison to exact retraining on the retain set or membership-inference/influence-function verification is described, which is load-bearing for establishing that targeted points have been removed.

minor comments (2)

Clarify the exact definition and computation of the 'forgetting quality' metric, including any formulas or pseudocode.
Add statistical significance tests or error bars to the performance comparisons to support the 'close to original' claim.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for their detailed review and constructive comments on our manuscript. We address each major comment below and plan to make revisions to strengthen the paper.

read point-by-point responses

Referee: [Abstract] Abstract: The abstract reports positive results on two datasets but provides no quantitative numbers, error bars, baseline comparisons, or details on how forgetting quality was measured. This makes the central empirical claims impossible to assess from the given information.

Authors: We agree that the abstract would benefit from more specific details to allow readers to assess the claims. In the revised manuscript, we will update the abstract to include quantitative results from our experiments, such as the accuracy and F1 scores achieved, the speedup factors compared to retraining, and a brief mention of how forgetting quality was evaluated through performance similarity metrics. revision: yes
Referee: [Evaluation] Evaluation section: The reported metrics (accuracy/F1 similarity to the original model and faster runtime) are compatible with the model retaining influence from the forget set via unchanged splits or leaf statistics. No comparison to exact retraining on the retain set or membership-inference/influence-function verification is described, which is load-bearing for establishing that targeted points have been removed.

Authors: This is a valid concern regarding the strength of evidence for actual unlearning. Our current approach evaluates by showing that the unlearned model maintains similar performance to the original while being faster than full retraining. To address this, we will add explicit comparisons to models retrained on the retain set only. For membership inference or influence function-based verification, these methods are not standard for XGBoost and tabular data; we will include a discussion of this limitation and why our metrics provide supporting evidence in this context. revision: partial

Circularity Check

0 steps flagged

No derivation chain present; purely empirical evaluation

full rationale

The manuscript is an applied empirical study introducing and benchmarking the XGBoost-Forget method on two tabular network-intrusion datasets. It reports performance, runtime, and forgetting-quality metrics but contains no equations, derivations, or first-principles claims that could reduce to their own inputs. Consequently there are no load-bearing steps of the enumerated circularity kinds. The central claim rests on experimental comparisons rather than any self-referential mathematical construction.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

No mathematical derivations or new entities are described; the work rests on standard supervised learning assumptions and the unstated premise that the chosen metrics capture true forgetting.

pith-pipeline@v0.9.1-grok · 5668 in / 1029 out tokens · 14563 ms · 2026-06-26T21:14:34.106841+00:00 · methodology

discussion (0)

Reference graph

Works this paper leans on

27 extracted references · 24 canonical work pages

[1]

In: 2025 13th ISDFS (2025)

Araújo, A., Rodrigues, D., Leite, P., Gonçalves, J.: Network intrusion detection system based on multiple datasets: Machine learning approaches. In: 2025 13th ISDFS (2025). https://doi.org/10.1109/ISDFS65363.2025.11011909

work page doi:10.1109/isdfs65363.2025.11011909 2025
[2]

Adversary instantiation: Lower bounds for differentially private machine learning,

Bourtoule, L., Chandrasekaran, V., Choquette-Choo, C.A., Jia, H., Travers, A., Zhang, B., Lie, D., Papernot, N.: Machine unlearning. In: 2021 IEEE Symposium on Security and Privacy (2021). https://doi.org/10.1109/SP40001.2021.00019

work page doi:10.1109/sp40001.2021.00019 2021
[3]

In: Proceedings of the 38th International Conf

Brophy, J., Lowd, D.: Machine unlearning for random forests. In: Proceedings of the 38th International Conf. on Machine Learning. PMLR (2021)

2021
[4]

Computers (2025)

Cevallos, I.D., Benalcázar, M.E., Valdivieso Caraguay, A.L., Zea, J.A., Barona- López, L.I.: A systematic literature review of machine unlearning techniques in neural networks. Computers (2025). https://doi.org/10.3390/computers14040150

work page doi:10.3390/computers14040150 2025
[5]

IAENG Inter- national Journal of Computer Science (2025)

Chen, W., Almamy, D.F.: Xgboost-driven intrusion detection method: Integrating smote-based class imbalance mitigation and multi-phase learning. IAENG Inter- national Journal of Computer Science (2025)

2025
[6]

Chowdhury, S.B.R., Choromanski, K., Sehanobish, A., Dubey, A., Chaturvedi, S.: Towards scalable exact machine unlearning using parameter-efficient fine-tuning (2025),https://arxiv.org/abs/2406.16257

arXiv 2025
[7]

IoT-23: A labeled dataset with malicious and benign IoT network traffic,

Garcia, S., Parmisano, A., Erquiaga, M.J.: Iot-23: A labeled dataset with malicious and benign iot network traffic (2020). https://doi.org/10.5281/zenodo.4743746

work page doi:10.5281/zenodo.4743746 2020
[8]

Network intrusion datasets: a survey, limitations, and recommendations,

Goldschmidt, P., Chudá, D.: Network intrusion datasets: A sur- vey, limitations, and recommendations. Computers & Security (2025). https://doi.org/10.1016/j.cose.2025.104510

work page doi:10.1016/j.cose.2025.104510 2025
[9]

In: Proceedings of the 31st International Joint Conf

Hu, H., Salčić, Z., Dobbie, G., Chen, J., Sun, L., Zhang, X.: Membership inference via backdooring. In: Proceedings of the 31st International Joint Conf. on Artificial Intelligence (2022). https://doi.org/10.24963/ijcai.2022/532

work page doi:10.24963/ijcai.2022/532 2022
[10]

Journal of Intelligent Systems (2024)

Issa, M., Aljanabi, M., Muhialdeen, H.: Systematic literature review on intrusion detectionsystems:Researchtrends,algorithms,methods,datasets,andlimitations. Journal of Intelligent Systems (2024). https://doi.org/10.1515/jisys-2023-0248

work page doi:10.1515/jisys-2023-0248 2024
[11]

Available: https://doi.org/10.1007/978-3-642-04898-2 616

Joyce, J.: Kullback-Leibler Divergence. Springer Berlin Heidelberg (2011). https://doi.org/10.1007/978-3-642-04898-2\_327

work page doi:10.1007/978-3-642-04898-2 2011
[12]

In: ICCST (2021)

Kiflay, A.Z., Tsokanos, A., Kirner, R.: A network intrusion detec- tion system using ensemble machine learning. In: ICCST (2021). https://doi.org/10.1109/ICCST49569.2021.9717397 12 D. Magalhães et al

work page doi:10.1109/iccst49569.2021.9717397 2021
[13]

In: Proceedings of the 29th ACM SIGKDD Conf

Lin, H., Chung, J.W., Lao, Y., Zhao, W.: Machine unlearning in gradient boosting decision trees. In: Proceedings of the 29th ACM SIGKDD Conf. on Knowledge Dis- covery and Data Mining. ACM (2023). https://doi.org/10.1145/3580305.3599420

work page doi:10.1145/3580305.3599420 2023
[14]

A Survey of Machine Unlearning

Nguyen,T.T.,Huynh,T.T.,Ren,Z.,Nguyen,P.L.,Liew,A.W.C.,Yin,H.,Nguyen, Q.V.H.: A survey of machine unlearning. ACM Trans. Intell. Syst. Technol. (2025). https://doi.org/10.1145/3749987

work page doi:10.1145/3749987 2025
[15]

In: 2022 International Conf

Nirupama, B.K., Niranjanamurthy, M.: Network intrusion detection using deci- sion tree and random forest. In: 2022 International Conf. on ACCAI (2022). https://doi.org/10.1109/ACCAI53970.2022.9752578

work page doi:10.1109/accai53970.2022.9752578 2022
[16]

In: 2023 InCACCT (2023)

Pandit, P.V., Bhushan, S., Waje, P.V.: Implementation of intrusion detection sys- tem using various machine learning approaches with ensemble learning. In: 2023 InCACCT (2023). https://doi.org/10.1109/InCACCT57535.2023.10141704

work page doi:10.1109/incacct57535.2023.10141704 2023
[17]

In: Proceedings of the 2024 European Interdisciplinary Cybersecurity Conf

Perifanis, V., Karypidis, E., Komodakis, N., Efraimidis, P.: Sftc: Machine unlearning via selective fine-tuning and targeted confusion. In: Proceedings of the 2024 European Interdisciplinary Cybersecurity Conf. ACM (2024). https://doi.org/10.1145/3655693.3655697

work page doi:10.1145/3655693.3655697 2024
[18]

In: Proceedings of the 11th Interna- tional Conf

Pinto, D., Vitorino, J., Maia, E., Amorim, I., Praça, I.: Flow exporter impact on intelligent intrusion detection systems. In: Proceedings of the 11th Interna- tional Conf. on Inf. Systems Security and Privacy - Volume 2: ICISSP (2025). https://doi.org/10.5220/0013131900003899

work page doi:10.5220/0013131900003899 2025
[19]

Computer (2024)

Qu, Y., Yuan, X., Ding, M., Ni, W., Rakotoarivelo, T., Smith, D.: Learn to unlearn: Insights into machine unlearning. Computer (2024). https://doi.org/10.1109/MC.2023.3333319

work page doi:10.1109/mc.2023.3333319 2024
[20]

In: 2024 IEEE 3rd International Conf

Rahman, M.S., Tausif Islam, W., Ahmed Khan, M.R.: Enhancing cy- bersecurity with an investigation into network intrusion detection sys- tem using machine learning. In: 2024 IEEE 3rd International Conf. on Robotics, Automation, Artificial-Intelligence and Internet-of-Things (2024). https://doi.org/10.1109/RAAICON64172.2024.10928505

work page doi:10.1109/raaicon64172.2024.10928505 2024
[21]

In: Proceedings of the 2021 International Conf

Schelter, S., Grafberger, S., Dunning, T.: Hedgecut: Maintaining randomised trees for low-latency machine unlearning. In: Proceedings of the 2021 International Conf. on Management of Data. ACM (2021). https://doi.org/10.1145/3448016.3457239

work page doi:10.1145/3448016.3457239 2021
[22]

Data in Brief (2025)

Silva, M., Pinto, D., Vitorino, J., Gonçalves, J., Maia, E., Praça, I.: Genis: A modular dataset for network intrusion detection and classification. Data in Brief (2025). https://doi.org/10.1016/j.dib.2025.111487

work page doi:10.1016/j.dib.2025.111487 2025
[23]

Frontiers in neurorobotics (2024)

Wang, C., Ying, Z., Pan, Z.: Machine unlearning in brain- inspired neural network paradigms. Frontiers in neurorobotics (2024). https://doi.org/10.3389/fnbot.2024.1361577

work page doi:10.3389/fnbot.2024.1361577 2024
[24]

Electronics (2024)

Wang, M., Yang, N., Guo, Y., Weng, N.: Learn-ids: Bridging gaps between datasets and learning-based network intrusion detection. Electronics (2024). https://doi.org/10.3390/electronics13061072

work page doi:10.3390/electronics13061072 2024
[25]

IEEE Transactions on Information Forensics and Security (2024)

Wang, W., Zhang, C., Tian, Z., Yu, S.: Machine unlearning via representation for- getting with parameter self-sharing. IEEE Transactions on Information Forensics and Security (2024). https://doi.org/10.1109/TIFS.2023.3331239

work page doi:10.1109/tifs.2023.3331239 2024
[26]

Wu, Z., Zhu, J., Li, Q., He, B.: Deltaboost: Gradient boosting decision trees with efficient machine unlearning. Proc. ACM Manag. Data (2023). https://doi.org/10.1145/3589313

work page doi:10.1145/3589313 2023
[27]

Machine Unlearning: Solutions and Challenges

Xu, J., Wu, Z., Wang, C., Jia, X.: Machine unlearning: Solutions and challenges. IEEE Transactions on Emerging Topics in Computational Intelligence (2024). https://doi.org/10.1109/TETCI.2024.3379240

work page doi:10.1109/tetci.2024.3379240 2024

[1] [1]

In: 2025 13th ISDFS (2025)

Araújo, A., Rodrigues, D., Leite, P., Gonçalves, J.: Network intrusion detection system based on multiple datasets: Machine learning approaches. In: 2025 13th ISDFS (2025). https://doi.org/10.1109/ISDFS65363.2025.11011909

work page doi:10.1109/isdfs65363.2025.11011909 2025

[2] [2]

Adversary instantiation: Lower bounds for differentially private machine learning,

Bourtoule, L., Chandrasekaran, V., Choquette-Choo, C.A., Jia, H., Travers, A., Zhang, B., Lie, D., Papernot, N.: Machine unlearning. In: 2021 IEEE Symposium on Security and Privacy (2021). https://doi.org/10.1109/SP40001.2021.00019

work page doi:10.1109/sp40001.2021.00019 2021

[3] [3]

In: Proceedings of the 38th International Conf

Brophy, J., Lowd, D.: Machine unlearning for random forests. In: Proceedings of the 38th International Conf. on Machine Learning. PMLR (2021)

2021

[4] [4]

Computers (2025)

Cevallos, I.D., Benalcázar, M.E., Valdivieso Caraguay, A.L., Zea, J.A., Barona- López, L.I.: A systematic literature review of machine unlearning techniques in neural networks. Computers (2025). https://doi.org/10.3390/computers14040150

work page doi:10.3390/computers14040150 2025

[5] [5]

IAENG Inter- national Journal of Computer Science (2025)

Chen, W., Almamy, D.F.: Xgboost-driven intrusion detection method: Integrating smote-based class imbalance mitigation and multi-phase learning. IAENG Inter- national Journal of Computer Science (2025)

2025

[6] [6]

Chowdhury, S.B.R., Choromanski, K., Sehanobish, A., Dubey, A., Chaturvedi, S.: Towards scalable exact machine unlearning using parameter-efficient fine-tuning (2025),https://arxiv.org/abs/2406.16257

arXiv 2025

[7] [7]

IoT-23: A labeled dataset with malicious and benign IoT network traffic,

Garcia, S., Parmisano, A., Erquiaga, M.J.: Iot-23: A labeled dataset with malicious and benign iot network traffic (2020). https://doi.org/10.5281/zenodo.4743746

work page doi:10.5281/zenodo.4743746 2020

[8] [8]

Network intrusion datasets: a survey, limitations, and recommendations,

Goldschmidt, P., Chudá, D.: Network intrusion datasets: A sur- vey, limitations, and recommendations. Computers & Security (2025). https://doi.org/10.1016/j.cose.2025.104510

work page doi:10.1016/j.cose.2025.104510 2025

[9] [9]

In: Proceedings of the 31st International Joint Conf

Hu, H., Salčić, Z., Dobbie, G., Chen, J., Sun, L., Zhang, X.: Membership inference via backdooring. In: Proceedings of the 31st International Joint Conf. on Artificial Intelligence (2022). https://doi.org/10.24963/ijcai.2022/532

work page doi:10.24963/ijcai.2022/532 2022

[10] [10]

Journal of Intelligent Systems (2024)

Issa, M., Aljanabi, M., Muhialdeen, H.: Systematic literature review on intrusion detectionsystems:Researchtrends,algorithms,methods,datasets,andlimitations. Journal of Intelligent Systems (2024). https://doi.org/10.1515/jisys-2023-0248

work page doi:10.1515/jisys-2023-0248 2024

[11] [11]

Available: https://doi.org/10.1007/978-3-642-04898-2 616

Joyce, J.: Kullback-Leibler Divergence. Springer Berlin Heidelberg (2011). https://doi.org/10.1007/978-3-642-04898-2\_327

work page doi:10.1007/978-3-642-04898-2 2011

[12] [12]

In: ICCST (2021)

Kiflay, A.Z., Tsokanos, A., Kirner, R.: A network intrusion detec- tion system using ensemble machine learning. In: ICCST (2021). https://doi.org/10.1109/ICCST49569.2021.9717397 12 D. Magalhães et al

work page doi:10.1109/iccst49569.2021.9717397 2021

[13] [13]

In: Proceedings of the 29th ACM SIGKDD Conf

Lin, H., Chung, J.W., Lao, Y., Zhao, W.: Machine unlearning in gradient boosting decision trees. In: Proceedings of the 29th ACM SIGKDD Conf. on Knowledge Dis- covery and Data Mining. ACM (2023). https://doi.org/10.1145/3580305.3599420

work page doi:10.1145/3580305.3599420 2023

[14] [14]

A Survey of Machine Unlearning

Nguyen,T.T.,Huynh,T.T.,Ren,Z.,Nguyen,P.L.,Liew,A.W.C.,Yin,H.,Nguyen, Q.V.H.: A survey of machine unlearning. ACM Trans. Intell. Syst. Technol. (2025). https://doi.org/10.1145/3749987

work page doi:10.1145/3749987 2025

[15] [15]

In: 2022 International Conf

Nirupama, B.K., Niranjanamurthy, M.: Network intrusion detection using deci- sion tree and random forest. In: 2022 International Conf. on ACCAI (2022). https://doi.org/10.1109/ACCAI53970.2022.9752578

work page doi:10.1109/accai53970.2022.9752578 2022

[16] [16]

In: 2023 InCACCT (2023)

Pandit, P.V., Bhushan, S., Waje, P.V.: Implementation of intrusion detection sys- tem using various machine learning approaches with ensemble learning. In: 2023 InCACCT (2023). https://doi.org/10.1109/InCACCT57535.2023.10141704

work page doi:10.1109/incacct57535.2023.10141704 2023

[17] [17]

In: Proceedings of the 2024 European Interdisciplinary Cybersecurity Conf

Perifanis, V., Karypidis, E., Komodakis, N., Efraimidis, P.: Sftc: Machine unlearning via selective fine-tuning and targeted confusion. In: Proceedings of the 2024 European Interdisciplinary Cybersecurity Conf. ACM (2024). https://doi.org/10.1145/3655693.3655697

work page doi:10.1145/3655693.3655697 2024

[18] [18]

In: Proceedings of the 11th Interna- tional Conf

Pinto, D., Vitorino, J., Maia, E., Amorim, I., Praça, I.: Flow exporter impact on intelligent intrusion detection systems. In: Proceedings of the 11th Interna- tional Conf. on Inf. Systems Security and Privacy - Volume 2: ICISSP (2025). https://doi.org/10.5220/0013131900003899

work page doi:10.5220/0013131900003899 2025

[19] [19]

Computer (2024)

Qu, Y., Yuan, X., Ding, M., Ni, W., Rakotoarivelo, T., Smith, D.: Learn to unlearn: Insights into machine unlearning. Computer (2024). https://doi.org/10.1109/MC.2023.3333319

work page doi:10.1109/mc.2023.3333319 2024

[20] [20]

In: 2024 IEEE 3rd International Conf

Rahman, M.S., Tausif Islam, W., Ahmed Khan, M.R.: Enhancing cy- bersecurity with an investigation into network intrusion detection sys- tem using machine learning. In: 2024 IEEE 3rd International Conf. on Robotics, Automation, Artificial-Intelligence and Internet-of-Things (2024). https://doi.org/10.1109/RAAICON64172.2024.10928505

work page doi:10.1109/raaicon64172.2024.10928505 2024

[21] [21]

In: Proceedings of the 2021 International Conf

Schelter, S., Grafberger, S., Dunning, T.: Hedgecut: Maintaining randomised trees for low-latency machine unlearning. In: Proceedings of the 2021 International Conf. on Management of Data. ACM (2021). https://doi.org/10.1145/3448016.3457239

work page doi:10.1145/3448016.3457239 2021

[22] [22]

Data in Brief (2025)

Silva, M., Pinto, D., Vitorino, J., Gonçalves, J., Maia, E., Praça, I.: Genis: A modular dataset for network intrusion detection and classification. Data in Brief (2025). https://doi.org/10.1016/j.dib.2025.111487

work page doi:10.1016/j.dib.2025.111487 2025

[23] [23]

Frontiers in neurorobotics (2024)

Wang, C., Ying, Z., Pan, Z.: Machine unlearning in brain- inspired neural network paradigms. Frontiers in neurorobotics (2024). https://doi.org/10.3389/fnbot.2024.1361577

work page doi:10.3389/fnbot.2024.1361577 2024

[24] [24]

Electronics (2024)

Wang, M., Yang, N., Guo, Y., Weng, N.: Learn-ids: Bridging gaps between datasets and learning-based network intrusion detection. Electronics (2024). https://doi.org/10.3390/electronics13061072

work page doi:10.3390/electronics13061072 2024

[25] [25]

IEEE Transactions on Information Forensics and Security (2024)

Wang, W., Zhang, C., Tian, Z., Yu, S.: Machine unlearning via representation for- getting with parameter self-sharing. IEEE Transactions on Information Forensics and Security (2024). https://doi.org/10.1109/TIFS.2023.3331239

work page doi:10.1109/tifs.2023.3331239 2024

[26] [26]

Wu, Z., Zhu, J., Li, Q., He, B.: Deltaboost: Gradient boosting decision trees with efficient machine unlearning. Proc. ACM Manag. Data (2023). https://doi.org/10.1145/3589313

work page doi:10.1145/3589313 2023

[27] [27]

Machine Unlearning: Solutions and Challenges

Xu, J., Wu, Z., Wang, C., Jia, X.: Machine unlearning: Solutions and challenges. IEEE Transactions on Emerging Topics in Computational Intelligence (2024). https://doi.org/10.1109/TETCI.2024.3379240

work page doi:10.1109/tetci.2024.3379240 2024