Generative Models for Personalized Federated Learning via Class-Separable Latent Spaces and Meta-Adaptation
Abstract
In decentralized learning scenarios such as federated learning (FL), client data are often highly heterogeneous, leading to non-IID distributions that degrade global model performance, slow convergence, and limit personalization. Conventional FL aggregation schemes inadequately capture client-specific structure and struggle to generalize across diverse data regimes. In this work, we propose a personalized federated learning framework that integrates Fisher Discriminant Analysis (FDA), meta-learning, and convolutional neural networks (CNNs) to jointly enhance class separability, adaptability, and representation quality. FDA is employed to project client-side features into an optimal class-separable latent space, enabling more discriminative modeling of local task characteristics under non-IID conditions.
On top of this representation, a meta-learning strategy is used to learn initialization parameters that support fast adaptation, allowing each client to efficiently fine-tune the model to its own data while preserving global knowledge. A CNN backbone is adopted to extract hierarchical feature representations, providing both fine-grained local patterns and robust global semantic embeddings suitable for heterogeneous federated environments. We conduct extensive
experiments on CIFAR-10, CIFAR-100, and EMNIST-Writers under non-IID data partitions generated using a Dirichlet concentration parameter α = 0.5. The proposed method achieves 79.6% test accuracy on CIFAR-10, 45.7% on CIFAR-100, and 85.5% on EMNIST-Writers, consistently outperforming strong baseline FL methods in terms of adaptability, generalization, and personalization quality. The results demonstrate that combining class-separable latent spaces with meta-adaptation improves both global coherence and client-specific performance while preserving data privacy and offering good scalability. Future work will extend the framework to multi-modal and sequential data and investigate its deployment in cross-device and cross-silo FL settings.
Keywords: federated learning; personalized models; Fisher Discriminant Analysis; meta-learning; non-IID data; data privacy.
Full Text:
PDFReferences
CHOWDHURY O., RISHAT M. A. S. A., AZAM M. H. B., and AMIN M. A. The rise of blockchain technology in Shariah based banking system. Proceedings of the 2nd International Conference on Computing Advancements, 2022: 349–358. https://doi.org/10.1145/3542954.3543005
XU Y., LIAO Y., WANG L., XU H., JIANG Z., and ZHANG W. Overcoming noisy labels and non-iid data in edge federated learning. IEEE Transactions on Mobile Computing, 2024. https://doi.org/10.1109/TMC.2024.3398801
HUANG G., WU Q., SUN P., MA Q., and CHEN X. Collaboration in federated learning with differential privacy: A stackelberg game analysis. IEEE Transactions on Parallel and Distributed Systems, 2024, 35(3): 455–469. https://doi.org/10.1109/TPDS.2024.3354713
BABAR M., QURESHI B., and KOUBAA A. Review on federated learning for digital transformation in healthcare through big data analytics. Future Generation Computer Systems, 2024. https://doi.org/10.1016/j.future.2024.05.046
QI Y., FENG Y., WANG X., LI H., and TIAN J. Leveraging federated learning and edge computing for recommendation systems within cloud computing networks. arXiv, 2024. https://doi.org/10.48550/arXiv.2403.03165
CHOWDHURY O., RISHAT M. A. S. A., AL-AMIN M., and AZAM M. H. B. The decentralized Shariah-based banking system in Bangladesh using blockchain technology. International Journal of Information Engineering and Electronic Business, 2023, 15(3): 12–28. https://doi.org/10.5815/ijieeb.2023.03.02
QUADIR B., MOSTAFA K., YANG J. C., SHEN J., and AKTER R. ARCS approach to PTA-based programming language practice sessions: Factors influencing programming problem-solving skills. Education and Information Technologies, 2023, 28(10): 13713–13735. https://doi.org/10.1007/s10639-023-11740-6
ROKAYA A., ISLAM S. M. T., ZHANG H., SUN L., ZHU M., and ZHAO L. Acceptance of chatbot based on emotional intelligence through machine learning algorithm. ICFEICT 2022, 2022: 610–616. https://doi.org/10.1109/ICFEICT57213.2022.00111
LU Z., PAN H., DAI Y., SI X., and ZHANG Y. Federated learning with non-iid data: A survey. IEEE Internet of Things Journal, 2024. https://doi.org/10.1109/JIOT.2024.3376548
WU X., PEI J., HAN X. H., CHEN Y. W., YAO J., LIU Y., et al. Fedel: Federated ensemble learning for non-iid data. Expert Systems with Applications, 2024, 237: 121390. https://doi.org/10.1016/j.eswa.2023.121390
ARIVAZHAGAN M. G., AGGARWAL V., SINGH A. K., and CHOUDHARY S. Federated learning with personalization layers. arXiv, 2019. https://doi.org/10.48550/arXiv.1912.00818
GHARI P. M. and SHEN Y. Personalized federated learning with mixture of models for adaptive prediction and model fine-tuning. arXiv, 2024. https://doi.org/10.48550/arXiv.2410.21547
ZHOU T., ZHANG J., and TSANG D. H. Fedfa: Federated learning with feature anchors to align features and classifiers for heterogeneous data. IEEE Transactions on Mobile Computing, 2023, 23(6): 6731–6742. https://doi.org/10.1109/TMC.2023.3325366
YE Z., ZHANG X., CHEN X., XIONG H., and YU D. Adaptive clustering based personalized federated learning framework for next POI recommendation with location noise. IEEE Transactions on Knowledge and Data Engineering, 2023, 36(5): 1843–1856. https://doi.org/10.1109/TKDE.2023.3312511
CHIEN W. C., LIN C. H., ZHU T., DAI C., GARG S., and MUKHERJEE A. Density-clustering aggregation for personalized federated learning with AI-enabled aerial and edge computing in UAVs. IEEE Internet of Things Journal, 2025. https://doi.org/10.1109/JIOT.2024.3522164
TAN A. Z., YU H., CUI L., and YANG Q. Towards personalized federated learning. IEEE Transactions on Neural Networks and Learning Systems, 2022, 34(12): 9587–9603. https://doi.org/10.1109/TNNLS.2022.3160699
ZHAO Y., PAN S., ZHAO Y., LIAO H., YE L., and ZHENG Y. Ultra-short-term wind power forecasting based on personalized robust federated learning with spatial collaboration. Energy, 2024, 288: 129847. https://doi.org/10.1016/j.energy.2023.129847
MCLAUGHLIN C. and SU L. Personalized federated learning via feature distribution adaptation. NeurIPS, 2024, 37: 77038–77059. https://doi.org/10.52202/079017-2451
GENG D., WANG S., and ZHANG Y. Multi-objective federated averaging algorithm. Expert Systems, 2025, 42(2): e13761. https://doi.org/10.1111/exsy.13761
KARAMI M. and KARAMI A. Harmony in federated learning: A comprehensive review of techniques to tackle heterogeneity and non-IID data. Cluster Computing, 2025 (in press). https://doi.org/10.1007/s10586-025-05250-y
THOMAS S. G. and MYAKALA P. K. Beyond the cloud: Federated learning and edge AI for the next decade. Journal of Computer and Communications, 2025, 13(2): 37–50. https://doi.org/10.4236/jcc.2025.132004
ULLAH I., DENG X., PEI X., MUSHTAQ H., and UZAIR M. IoV-SFL: A blockchain-based federated learning framework for secure and efficient data sharing in the internet of vehicles. Peer-to-Peer Networking and Applications, 2025, 18(1): 1–20. https://doi.org/10.1007/s12083-024-01821-9
LI R., WANG H., LU Q., YAN J., JI S., and MA Y. Research on medical image classification based on improved FedAvg algorithm. Tsinghua Science and Technology, 2025. https://doi.org/10.26599/TST.2024.9010184
ZHANG X., HONG M., DHOPLE S., YIN W., and LIU Y. FedPD: A federated learning framework with adaptivity to non-IID data. IEEE Transactions on Signal Processing, 2021, 69: 6055–6070. https://doi.org/10.1109/TSP.2021.3115952
DORIGUZZI-CORIN R. and SIRACUSA D. FLAD: Adaptive federated learning for DDoS attack detection. Computers & Security, 2024, 137: 103597. https://doi.org/10.1016/j.cose.2023.103597
ZENG Y., MU Y., YUAN J., TENG S., ZHANG J., WAN J., et al. Adaptive federated learning with non-IID data. The Computer Journal, 2023, 66(11): 2758–2772. https://doi.org/10.1093/comjnl/bxac118
AMIN M. S., KIM S., RISHAT M. A. S. A., TANG Z., and AHN H. A systematic literature review of privacy information disclosure in AI-integrated IoT technologies. Sustainability, 2024, 17(1): 8. https://doi.org/10.3390/su17010008
ZHANG J., HUA Y., WANG H., SONG T., XUE Z., MA R., et al. GPFL: Simultaneously learning global and personalized feature information for personalized federated learning. ICCV, 2023: 5041–5051. https://doi.org/10.48550/arXiv.2308.10279
JANG J. and CHOI B. J. Personalized federated learning via deviation tracking representation learning. ICOIN 2024, 2024: 762–766. https://doi.org/10.1109/ICOIN59985.2024.10572208
SHEN W., WU S., and TAO Y. CLDP-pFedAvg: Safeguarding client data privacy in personalized federated averaging. Mathematics, 2024, 12(22): 3630. https://doi.org/10.3390/math12223630
XU J., TONG X., and HUANG S. L. Personalized federated learning with feature alignment and classifier collaboration. arXiv, 2023. https://doi.org/10.48550/arXiv.2306.11867
WANG J., LIU Q., LIANG H., JOSHI G., and POOR H. V. Tackling the objective inconsistency problem in heterogeneous federated optimization. NeurIPS, 2020, 33: 7611–7623. https://doi.org/10.48550/arXiv.2007.07481
MCMAHAN B., MOORE E., RAMAGE D., HAMPSON S., and Y ARCAS B. A. Communication-efficient learning of deep networks from decentralized data. AISTATS, 2017: 1273–1282. https://doi.org/10.48550/arXiv.1602.05629
KHAN U., PAO W., PILARIO K. E. S., SALLIH N., and KHAN M. R. Two-phase flow regime identification using multi-method feature extraction and explainable kernel Fisher discriminant analysis. International Journal of Numerical Methods for Heat & Fluid Flow, 2024, 34(8): 2836–2864. https://doi.org/10.1108/HFF-09-2023-0526
BACCOUR E., ERBAD A., MOHAMED A., HAMDI M., and GUIZANI M. A blockchain-based reliable federated meta-learning for metaverse: A dual game framework. IEEE Internet of Things Journal, 2024. https://doi.org/10.1109/JIOT.2024.3383096
Refbacks
- There are currently no refbacks.
