pith. sign in

arxiv: 2507.07316 · v1 · pith:2Z67CUYRnew · submitted 2025-07-09 · 💻 cs.LG · cs.CR

AdeptHEQ-FL: Adaptive Homomorphic Encryption for Federated Learning of Hybrid Classical-Quantum Models with Dynamic Layer Sparing

Md Abrar Jahin , Taufikur Rahman Fuad , M. F. Mridha , Nafiz Fahad , Md. Jakir Hossen This is my paper

Pith reviewed 2026-05-19 05:04 UTC · model grok-4.3

classification 💻 cs.LG cs.CR
keywords Federated LearningHomomorphic EncryptionHybrid Quantum-Classical ModelsPrivacy PreservationCommunication EfficiencyNon-IID DataDynamic Layer Freezing
0
0 comments X

The pith

AdeptHEQ-FL combines hybrid CNN-PQC models with selective homomorphic encryption and dynamic layer freezing to raise accuracy and cut communication costs in privacy-preserving federated learning.

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

The paper presents AdeptHEQ-FL as a unified framework for federated learning that pairs classical convolutional networks with quantum parameterised circuits. It adds an accuracy-weighted aggregation step that uses differentially private validation scores, applies homomorphic encryption only to selected layers, and freezes less important layers on the fly to lower communication volume. The design targets non-IID data distributions while keeping formal privacy guarantees and convergence properties. Experiments on CIFAR-10, SVHN, and Fashion-MNIST report accuracy lifts of roughly 25 percent over a standard federated quantum baseline and 14 percent over a fully homomorphically encrypted version. The approach keeps quantum layers adaptable even after freezing decisions are made.

Core claim

AdeptHEQ-FL achieves improved accuracy and lower communication overhead in hybrid classical-quantum federated learning by combining a CNN-PQC architecture, adaptive accuracy-weighted aggregation, selective homomorphic encryption on sensitive layers, and dynamic layer-wise freezing, while supplying formal privacy guarantees and convergence analysis.

What carries the argument

Hybrid CNN-PQC architecture with selective homomorphic encryption and dynamic layer-wise adaptive freezing for secure, low-overhead aggregation.

If this is right

  • Accuracy gains of approximately 25 percent over standard federated quantum networks on CIFAR-10 follow from the combined mechanisms.
  • Communication overhead decreases by freezing less important layers while accuracy is maintained.
  • Formal privacy guarantees hold because encryption is applied selectively and aggregation uses differentially private accuracies.
  • Quantum adaptability remains intact despite layer freezing decisions.

Where Pith is reading between the lines

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

  • The selective-encryption pattern could transfer to purely classical federated learning to reduce encryption costs.
  • Extending the dynamic freezing rule to other quantum circuit depths might further lower resource use on different hardware.
  • Testing the framework on sequential decision tasks rather than image classification would reveal whether the accuracy-communication trade-off generalises.

Load-bearing premise

The hybrid model and selective freezing choices preserve quantum adaptability and convergence without introducing bias from layer selection or encryption in non-IID settings.

What would settle it

Running the same non-IID CIFAR-10 experiments without dynamic layer freezing and checking whether communication volume stays high or accuracy falls back to baseline levels would test the central claim.

Figures

Figures reproduced from arXiv: 2507.07316 by Md Abrar Jahin, Md. Jakir Hossen, M. F. Mridha, Nafiz Fahad, Taufikur Rahman Fuad.

Figure 1
Figure 1. Figure 1: This flowchart provides an overview of the AdeptHEQ-FL framework and illustrates the multi-stage process, detailing the [PITH_FULL_IMAGE:figures/full_fig_p004_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: 4-qubit 2-layered PQC of AdeptHEQ-FL comprising amplitude embedding, two Strongly Entangling Layers (parameterized [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
read the original abstract

Federated Learning (FL) faces inherent challenges in balancing model performance, privacy preservation, and communication efficiency, especially in non-IID decentralized environments. Recent approaches either sacrifice formal privacy guarantees, incur high overheads, or overlook quantum-enhanced expressivity. We introduce AdeptHEQ-FL, a unified hybrid classical-quantum FL framework that integrates (i) a hybrid CNN-PQC architecture for expressive decentralized learning, (ii) an adaptive accuracy-weighted aggregation scheme leveraging differentially private validation accuracies, (iii) selective homomorphic encryption (HE) for secure aggregation of sensitive model layers, and (iv) dynamic layer-wise adaptive freezing to minimize communication overhead while preserving quantum adaptability. We establish formal privacy guarantees, provide convergence analysis, and conduct extensive experiments on the CIFAR-10, SVHN, and Fashion-MNIST datasets. AdeptHEQ-FL achieves a $\approx 25.43\%$ and $\approx 14.17\%$ accuracy improvement over Standard-FedQNN and FHE-FedQNN, respectively, on the CIFAR-10 dataset. Additionally, it reduces communication overhead by freezing less important layers, demonstrating the efficiency and practicality of our privacy-preserving, resource-aware design for FL.

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.

Referee Report

1 major / 2 minor

Summary. The manuscript introduces AdeptHEQ-FL, a hybrid classical-quantum federated learning framework that combines a CNN-PQC architecture, an adaptive accuracy-weighted aggregation scheme using differentially private validation accuracies, selective homomorphic encryption for sensitive layers, and dynamic layer-wise freezing to reduce communication overhead. The authors claim formal privacy guarantees, provide a convergence analysis, and report accuracy gains of approximately 25.43% over Standard-FedQNN and 14.17% over FHE-FedQNN on CIFAR-10, with efficiency improvements demonstrated on CIFAR-10, SVHN, and Fashion-MNIST.

Significance. If the reported accuracy improvements and convergence properties hold after accounting for the adaptive mechanisms, the work could advance privacy-preserving and resource-efficient quantum-enhanced federated learning in non-IID settings. The combination of selective encryption and dynamic freezing addresses practical FL challenges, though empirical and theoretical support requires strengthening for the claims to be fully convincing.

major comments (1)
  1. [Convergence Analysis] Convergence Analysis section: The analysis appears to assume fixed full-model updates with uniform participation (standard FL setup). However, the dynamic layer-wise freezing and selective HE (detailed in the method) produce time-varying partial aggregation that is not re-derived for the non-IID case. This directly threatens the non-IID convergence guarantee that supports the practicality claims.
minor comments (2)
  1. [Abstract] Abstract: Reports specific accuracy numbers (≈25.43% and ≈14.17%) without error bars, dataset splits, ablation details, or statistical tests, limiting verification of the central performance claims.
  2. [Method] Method description: The accuracy-weighted aggregation and layer freezing threshold are introduced as free parameters; their selection process and potential impact on quantum adaptability or introduction of biases should be clarified with additional experiments.

Simulated Author's Rebuttal

1 responses · 0 unresolved

We thank the referee for the constructive comments on our manuscript. We address the major comment on the convergence analysis below and will strengthen the theoretical section accordingly.

read point-by-point responses

  1. Referee: [Convergence Analysis] Convergence Analysis section: The analysis appears to assume fixed full-model updates with uniform participation (standard FL setup). However, the dynamic layer-wise freezing and selective HE (detailed in the method) produce time-varying partial aggregation that is not re-derived for the non-IID case. This directly threatens the non-IID convergence guarantee that supports the practicality claims.

    Authors: We acknowledge that the current convergence analysis is presented under standard FL assumptions with full-model updates to establish the base result. The dynamic layer-wise freezing and selective HE indeed introduce time-varying partial updates that are not explicitly re-derived for the non-IID setting. We will revise the Convergence Analysis section to model layer freezing as a dynamic masking operator on the update vector and incorporate the accuracy-weighted aggregation into the bound. The extended analysis will add terms for the effective sparsity and heterogeneity mitigation, yielding a non-IID convergence guarantee that accounts for these mechanisms while preserving the original privacy and efficiency claims. revision: yes

Circularity Check

0 steps flagged

No circularity: claims rest on experimental results and stated formal analysis without self-referential reduction

full rationale

The paper presents a hybrid CNN-PQC FL framework with adaptive aggregation, selective HE, and dynamic freezing. It reports empirical accuracy gains on CIFAR-10 (≈25.43% and ≈14.17% over baselines) and states that formal privacy guarantees plus convergence analysis are provided. No equations or sections in the supplied text define a quantity in terms of itself, rename a fitted parameter as a prediction, or rely on a load-bearing self-citation whose content reduces to the present result. The convergence claim is presented as an independent derivation rather than a re-labeling of the adaptive mechanisms; therefore the central results do not collapse to their inputs by construction.

Axiom & Free-Parameter Ledger

2 free parameters · 2 axioms · 0 invented entities

Abstract-only review limits visibility into exact parameters; design choices such as layer importance scoring and encryption selectivity are introduced without upstream justification or independent evidence.

free parameters (2)
  • accuracy-weighted aggregation weights
    Derived from differentially private validation accuracies; likely tuned per round or dataset.
  • layer freezing threshold
    Controls dynamic sparing; value chosen to balance overhead and adaptability.
axioms (2)
  • domain assumption Hybrid CNN-PQC architecture supplies greater expressivity than classical-only models in decentralized settings
    Invoked to justify the core model choice without proof in the abstract.
  • domain assumption Selective homomorphic encryption on sensitive layers preserves formal privacy guarantees
    Stated as established but location and proof details absent from abstract.

pith-pipeline@v0.9.0 · 5777 in / 1421 out tokens · 35311 ms · 2026-05-19T05:04:12.429406+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

What do these tags mean?
matches
The paper's claim is directly supported by a theorem in the formal canon.
supports
The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends
The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses
The paper appears to rely on the theorem as machinery.
contradicts
The paper's claim conflicts with a theorem or certificate in the canon.
unclear
Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Reference graph

Works this paper leans on

42 extracted references · 42 canonical work pages · 2 internal anchors

  1. [1]

    Exploring Homomorphic Encryption and Differential Privacy Techniques towards Secure Federated Learning Paradigm

    Rezak Aziz, Soumya Banerjee, Samia Bouzefrane, and Thinh Le Vinh. Exploring Homomorphic Encryption and Differential Privacy Techniques towards Secure Federated Learning Paradigm. Future Internet, 15(9):310, 2023. 2

    work page 2023

  2. [2]

    Pri- vacy Preserving Federated Learning: A Novel Approach for Combining Differential Privacy and Homomorphic Encryp- tion

    Rezak Aziz, Soumya Banerjee, and Samia Bouzefrane. Pri- vacy Preserving Federated Learning: A Novel Approach for Combining Differential Privacy and Homomorphic Encryp- tion. In Information Security Theory and Practice - 14th IFIP WG 11.2 International Conference, WISTP 2024, Paris, France, February 29 - March 1, 2024, Proceedings , pages 162–177. Springer...

    work page 2024

  3. [3]

    Ville Bergholm, Josh Izaac, Maria Schuld, Christian Gogolin, M. Sohaib Alam, Shahnawaz Ahmed, Juan Miguel Arrazola, Carsten Blank, Alain Delgado, Soran Jahangiri, Keri McK- iernan, Johannes Jakob Meyer, Zeyue Niu, Antal Száva, and Nathan Killoran. PennyLane: Automatic differentiation of hybrid quantum-classical computations, 2018. 7

    work page 2018

  4. [4]

    Application of quantum-inspired tensor networks to optimize federated learning systems

    Amandeep Singh Bhatia, Mandeep Kaur Saggi, and Sabre Kais. Application of quantum-inspired tensor networks to optimize federated learning systems. Quantum Mach. Intell., 7(1):12, 2025. 2

    work page 2025

  5. [5]

    Privacy-Preserving in Federated Learning: A Comparison between Differential Privacy and Homomorphic Encryption across Different Scenarios

    Alessio Catalfamo, Maria Fazio, Antonio Celesti, and Mas- simo Villari. Privacy-Preserving in Federated Learning: A Comparison between Differential Privacy and Homomorphic Encryption across Different Scenarios. In IEEE International Conference on Software Testing, Verification and Validation, ICST 2025 - Workshops, Naples, Italy, March 31 - April 4, 2025,...

    work page 2025

  6. [6]

    Foundations of Quantum Federated Learning Over Classical and Quantum Networks

    Mahdi Chehimi, Samuel Yen-Chi Chen, Walid Saad, Don Towsley, and Mérouane Debbah. Foundations of Quantum Federated Learning Over Classical and Quantum Networks. IEEE Netw., 38(1):124–130, 2024. 2

    work page 2024

  7. [7]

    Se- cure and efficient federated learning via novel multi-party computation and compressed sensing

    Lvjun Chen, Di Xiao, Zhuyang Yu, and Maolan Zhang. Se- cure and efficient federated learning via novel multi-party computation and compressed sensing. Inf. Sci., 667:120481,

    work page

  8. [8]

    Federated Learning with Privacy Preservation in Large-Scale Distributed Systems Using Differential Privacy and Homomorphic Encryption

    Yue Chen, Yufei Yang, Yingwei Liang, Taipeng Zhu, and Dehui Huang. Federated Learning with Privacy Preservation in Large-Scale Distributed Systems Using Differential Privacy and Homomorphic Encryption. Informatica (Slovenia), 49 (13), 2025. 2

    work page 2025

  9. [9]

    Homomorphic Encryption for Arithmetic of Approx- imate Numbers

    Jung Hee Cheon, Andrey Kim, Miran Kim, and Yong Soo Song. Homomorphic Encryption for Arithmetic of Approx- imate Numbers. In Advances in Cryptology - ASIACRYPT 2017 - 23rd International Conference on the Theory and Applications of Cryptology and Information Security, Hong Kong, China, December 3-7, 2017, Proceedings, Part I, pages 409–437. Springer, 2017. 6, 7

    work page 2017

  10. [10]

    Karanth, Pedro Maciel Xavier, Iago Leal de Freitas, Nouhaila Innan, Sadok Ben Yahia, Muhammad Shafique, and David E

    Siddhant Dutta, Pavana P. Karanth, Pedro Maciel Xavier, Iago Leal de Freitas, Nouhaila Innan, Sadok Ben Yahia, Muhammad Shafique, and David E. Bernal Neira. Feder- ated Learning with Quantum Computing and Fully Homo- morphic Encryption: A Novel Computing Paradigm Shift in Privacy-Preserving ML. CoRR, abs/2409.11430, 2024. arXiv: 2409.11430. 1, 2, 7, 8

    work page arXiv 2024

  11. [11]

    The Algorithmic Founda- tions of Differential Privacy

    Cynthia Dwork and Aaron Roth. The Algorithmic Founda- tions of Differential Privacy. Foundations and Trends® in Theoretical Computer Science, 9(3-4):211–407, 2013. Pub- lisher: Now Publishers. 6

    work page 2013

  12. [12]

    Deep Learning

    Ian Goodfellow, Yoshua Bengio, and Aaron Courville. Deep Learning . MIT Press, 2016. http : / / www . deeplearningbook.org. 6

    work page 2016

  13. [13]

    FedSIGN: A sign-based federated learning framework with privacy and robustness guarantees

    Zhenyuan Guo, Lei Xu, and Liehuang Zhu. FedSIGN: A sign-based federated learning framework with privacy and robustness guarantees. Comput. Secur., 135:103474, 2023. 1, 2

    work page 2023

  14. [14]

    Privacy-Preserving Hybrid Feder- ated Learning Framework for Mental Healthcare Applica- tions: Clustered and Quantum Approaches

    Arti Gupta, Manish Kumar Maurya, Khyati Dhere, and Vi- jay Kumar Chaurasiya. Privacy-Preserving Hybrid Feder- ated Learning Framework for Mental Healthcare Applica- tions: Clustered and Quantum Approaches. IEEE Access, 12: 145054–145068, 2024. 3

    work page 2024

  15. [15]

    Pcosgen-test dataset, 2024

    Palak Handa, Anushka Saini, Siddhant Dutta, Harsh Pathak, Nishi Choudhary, Nidhi Goel, and Jasdeep Kaur Dhanao. Pcosgen-test dataset, 2024. 2

    work page 2024

  16. [16]

    Array programming with NumPy

    Charles R Harris, K Jarrod Millman, Stéfan J Van Der Walt, Ralf Gommers, Pauli Virtanen, David Cournapeau, Eric Wieser, Julian Taylor, Sebastian Berg, Nathaniel J Smith, et al. Array programming with NumPy. Nature, 585(7825): 357–362, 2020. 7

    work page 2020

  17. [17]

    FedQNN: Federated Learning using Quantum Neural Networks

    Nouhaila Innan, Muhammad Al-Zafar Khan, Alberto Marchi- sio, Muhammad Shafique, and Mohamed Bennai. FedQNN: Federated Learning using Quantum Neural Networks. In International Joint Conference on Neural Networks, IJCNN 2024, Yokohama, Japan, June 30 - July 5, 2024, pages 1–9. IEEE, 2024. 1, 2

    work page 2024

  18. [18]

    Qamplifynet: pushing the boundaries of supply chain backo- rder prediction using interpretable hybrid quantum-classical neural network

    Md Abrar Jahin, Md Sakib Hossain Shovon, Md Saiful Islam, Jungpil Shin, Muhammad Firoz Mridha, and Yuichi Okuyama. Qamplifynet: pushing the boundaries of supply chain backo- rder prediction using interpretable hybrid quantum-classical neural network. Scientific Reports, 13(1):18246, 2023. 3

    work page 2023

  19. [19]

    Akmol Masud, Md Wahiduzzaman Suva, M

    Md Abrar Jahin, Md. Akmol Masud, Md Wahiduzzaman Suva, M. F. Mridha, and Nilanjan Dey. Lorentz-Equivariant Quantum Graph Neural Network for High-Energy Physics. IEEE Transactions on Artificial Intelligence , pages 1–11,

    work page

  20. [20]

    Brendan McMahan, Brendan Avent, Au- rélien Bellet, Mehdi Bennis, Arjun Nitin Bhagoji, Kallista A

    Peter Kairouz, H. Brendan McMahan, Brendan Avent, Au- rélien Bellet, Mehdi Bennis, Arjun Nitin Bhagoji, Kallista A. Bonawitz, Zachary Charles, Graham Cormode, Rachel Cum- mings, Rafael G. L. D’Oliveira, Hubert Eichner, Salim El Rouayheb, David Evans, Josh Gardner, Zachary Garrett, Adrià Gascón, Badih Ghazi, Phillip B. Gibbons, Marco Gruteser, Zaïd Harchao...

    work page 2021

  21. [21]

    Awaysheh, Sadi Alawadi, and Liina Kamm

    Hiroki Kaminaga, Feras M. Awaysheh, Sadi Alawadi, and Liina Kamm. MPCFL: Towards Multi-party Computation for Secure Federated Learning Aggregation. In Proceedings of the IEEE/ACM 16th International Conference on Utility and Cloud Computing, UCC 2023, Taormina (Messina), Italy, December 4-7, 2023, page 19. ACM, 2023. 2

    work page 2023

  22. [22]

    Kingma and Jimmy Ba

    Diederik P. Kingma and Jimmy Ba. Adam: A Method for Stochastic Optimization. In 3rd International Conference on Learning Representations, ICLR 2015, San Diego, CA, USA, May 7-9, 2015, Conference Track Proceedings, 2015. 6

    work page 2015

  23. [23]

    Learning multiple layers of features from tiny images

    Alex Krizhevsky and Geoffrey Hinton. Learning multiple layers of features from tiny images. Technical Report 0, University of Toronto, Toronto, Ontario, 2009. 2, 7, 8

    work page 2009

  24. [24]

    Experimentally validated quantum-secure federated learning over a multi-user quantum network

    Zhi-Ping Liu, Xiao-Yu Cao, Hao-Wen Liu, Xiao-Ran Sun, Yu Bao, Yu-Shuo Lu, Hua-Lei Yin, and Zeng-Bing Chen. Practi- cal quantum federated learning and its experimental demon- stration. CoRR, abs/2501.12709, 2025. arXiv: 2501.12709. 1, 2

    work page internal anchor Pith review arXiv 2025

  25. [25]

    Data Structures for Statistical Computing in Python, 2010

    Wes McKinney. Data Structures for Statistical Computing in Python, 2010. 7

    work page 2010

  26. [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 Proceedings of the 20th International Conference on Artificial Intelligence and Statistics , pages 1273–1282. PMLR, 2017. 1, 5

    work page 2017

  27. [27]

    Communication- Efficient Learning of Deep Networks from Decentralized Data

    Brendan McMahan, Eider Moore, Daniel Ramage, Seth Hampson, and Blaise Agüera y Arcas. Communication- Efficient Learning of Deep Networks from Decentralized Data. In Proceedings of the 20th International Conference on Artificial Intelligence and Statistics, AISTATS 2017, 20- 22 April 2017, Fort Lauderdale, FL, USA, pages 1273–1282. PMLR, 2017. 2

    work page 2017

  28. [28]

    Yuval Netzer, Tao Wang, Adam Coates, Alessandro Bissacco, Bo Wu, and Andrew Y . Ng. Reading digits in natural images with unsupervised feature learning. In NIPS Workshop on Deep Learning and Unsupervised Feature Learning 2011 ,

    work page 2011

  29. [29]

    Brain Tumor MRI Dataset, 2021

    Msoud Nickparvar. Brain Tumor MRI Dataset, 2021. 2

    work page 2021

  30. [30]

    TenSEAL: A library for doing Homomorphic Encryption operations on tensors, 2020

    OpenMined Community. TenSEAL: A library for doing Homomorphic Encryption operations on tensors, 2020. 7

    work page 2020

  31. [31]

    PyTorch: An Imperative Style, High-Performance Deep Learning Library

    Adam Paszke, Sam Gross, Francisco Massa, Adam Lerer, James Bradbury, Gregory Chanan, Trevor Killeen, Zeming Lin, Natalia Gimelshein, Luca Antiga, et al. PyTorch: An Imperative Style, High-Performance Deep Learning Library. In Advances in Neural Information Processing Systems 32 , pages 8024–8035, 2019. 7

    work page 2019

  32. [32]

    Federated Quantum Machine Learning with Differential Privacy

    Rod Rofougaran, Shinjae Yoo, Huan-Hsin Tseng, and Samuel Yen-Chi Chen. Federated Quantum Machine Learning with Differential Privacy. In IEEE International Conference on Acoustics, Speech and Signal Processing, ICASSP 2024, Seoul, Republic of Korea, April 14-19, 2024 , pages 9811–

    work page 2024

  33. [33]

    A generic framework for privacy preserving deep learning

    Théo Ryffel, Andrew Trask, Morten Dahl, Bobby Wagner, Jason E Mancuso, Daniel Rueckert, and Jonathan Passerat- Palmbach. A generic framework for privacy preserving deep learning. In NeurIPS Workshop on Privacy Preserving Ma- chine Learning, 2018. 7

    work page 2018

  34. [34]

    Adaptive Federated Learning with Functional Encryption: A Comparison of Classical and Quantum-safe Options

    Enrico Sorbera, Federica Zanetti, Giacomo Brandi, Alessan- dro Tomasi, Roberto Doriguzzi Corin, and Silvio Ranise. Adaptive Federated Learning with Functional Encryption: A Comparison of Classical and Quantum-safe Options. CoRR, abs/2504.00563, 2025. arXiv: 2504.00563. 2

    work page arXiv 2025

  35. [35]

    Combining homomorphic encryption and differential privacy in federated learning

    Arnaud Grivet Sébert, Marina Checri, Oana Stan, Renaud Sirdey, and Cédric Gouy-Pailler. Combining homomorphic encryption and differential privacy in federated learning. In 20th Annual International Conference on Privacy, Security and Trust, PST 2023, Copenhagen, Denmark, August 21-23, 2023, pages 1–7. IEEE, 2023. 2

    work page 2023

  36. [36]

    Quantum-Inspired Privacy-Preserving Federated Learning Framework for Secure Dementia Classification

    Gazi Tanbhir and Md Farhan Shahriyar. Quantum-Inspired Privacy-Preserving Federated Learning Framework for Secure Dementia Classification. CoRR, abs/2503.03267, 2025. arXiv: 2503.03267. 2

    work page arXiv 2025

  37. [37]

    A Novel Privacy-Preserving Federated Learning Model Based on Secure Multi-party Computation

    Anh-Tu Tran, The Dung Luong, and Xuan Sang Pham. A Novel Privacy-Preserving Federated Learning Model Based on Secure Multi-party Computation. In Integrated Uncer- tainty in Knowledge Modelling and Decision Making - 10th International Symposium, IUKM 2023, Kanazawa, Japan, November 2-4, 2023, Proceedings, Part II, pages 321–333. Springer, 2023. 2

    work page 2023

  38. [38]

    Quan- tum Enhanced Federated Learning with Differential Privacy

    Shoaib Ullah, Madam Hussain Shah, and Adeel Anjum. Quan- tum Enhanced Federated Learning with Differential Privacy. In International Conference on Frontiers of Information Tech- nology, FIT 2024, Islamabad, Pakistan, December 9-10, 2024, pages 1–6. IEEE, 2024. 1, 2

    work page 2024

  39. [39]

    ADPHE-FL: Federated learning method based on adaptive differential privacy and homomorphic encryption

    Tao Wu, Yulin Deng, Qizhao Zhou, Xi Chen, and Ming Zhang. ADPHE-FL: Federated learning method based on adaptive differential privacy and homomorphic encryption. Peer Peer Netw. Appl., 18(3):141, 2025. 2

    work page 2025

  40. [40]

    Fashion-MNIST: a Novel Image Dataset for Benchmarking Machine Learning Algorithms

    Han Xiao, Kashif Rasul, and Roland V ollgraf. Fashion- MNIST: A Novel Image Dataset for Benchmarking Machine Learning Algorithms. arXiv preprint arXiv:1708.07747, 2017. 7, 8

    work page internal anchor Pith review Pith/arXiv arXiv 2017

  41. [41]

    Towards Quantum-Safe Federated Learn- ing via Homomorphic Encryption: Learning with Gradients

    Guangfeng Yan, Shanxiang Lyu, Hanxu Hou, Zhiyong Zheng, and Linqi Song. Towards Quantum-Safe Federated Learn- ing via Homomorphic Encryption: Learning with Gradients. CoRR, abs/2402.01154, 2024. arXiv: 2402.01154. 1, 2

    work page arXiv 2024

  42. [42]

    Secure Federated Learning Scheme Based on Differential Privacy and Homo- morphic Encryption

    Xuyan Zhang, Da Huang, and Yuhua Tang. Secure Federated Learning Scheme Based on Differential Privacy and Homo- morphic Encryption. In Advanced Intelligent Computing Technology and Applications - 20th International Conference, ICIC 2024, Tianjin, China, August 5-8, 2024, Proceedings, Part V (LNAI), pages 435–446. Springer, 2024. 2

    work page 2024