Skip to main content

Resultados relacionados

Resultados totales (Incluyendo duplicados): 3
Encontrada(s) 1 página(s)

RUA. Repositorio Institucional de la Universidad de Alicante
Publication Otros (Other)

Privacy and Accuracy Implications of Model Complexity and Integration in Heterogeneous Federated Learning

  • Németh, Gergely Dániel
  • Lozano, Miguel Angel
  • Quadrianto, Novi
  • Oliver, Nuria
  • Lozano, Miguel Angel|||/items/413f111d-7a63-43f2-bb70-6421affe1647
Federated Learning (FL) has been proposed as a privacy-preserving solution for distributed machine learning, particularly in heterogeneous FL settings where clients have varying computational capabilities and thus train models with different complexities compared to the server’s model. However, FL is not without vulnerabilities: recent studies have shown that it is susceptible to membership inference attacks (MIA), which can compromise the privacy of client data. In this paper, we examine the intersection of these two aspects, heterogeneous FL and its privacy vulnerabilities, by focusing on the role of client model integration, the process through which the server integrates parameters from clients’ smaller models into its larger model. To better understand this process, we first propose a taxonomy that categorizes existing heterogeneous FL methods and enables the design of seven novel heterogeneous FL model integration strategies. Using CIFAR-10, CIFAR-100, and FEMNIST vision datasets, we evaluate the privacy and accuracy trade-offs of these approaches under three types of MIAs. Our findings reveal significant differences in privacy leakage and performance depending on the integration method. Notably, introducing randomness in the model integration process enhances client privacy while maintaining competitive accuracy for both the clients and the server. This work provides quantitative light on the privacy-accuracy implications client model integration in heterogeneous FL settings, paving the way towards more secure and efficient FL systems.
RUA. Repositorio Institucional de la Universidad de Alicante
Publication Otros (Other)

Relaxing Core Assumptions: the Impact of Data, Model and Participation Heterogeneity on Performance, Privacy and Fairness in Federated Learning

  • Németh, Gergely Dániel
Federated Learning (FL) enables decentralized training of machine learning models on distributed data while preserving privacy by design. An FL design consists of clients training models on private data and a central server aggregating a global model based on the consensus among clients. In an ideal scenario, the training data and computing resources are identically and independently distributed (i.i.d.) among clients, therefore, clients can work together in agreement to reach a global optima. However, in realistic FL settings, heterogeneity arises between clients in terms of both data and resource availability. This research focuses on such scenarios, with a special interest on how the server can adapt the aggregation method from a simple averaging to address the clients’ diversity. The first research direction discusses existing client selection methods and proposes a novel taxonomy of FL methods where the participation of the clients is actively managed by the server to achieve a global objective with respect to the client heterogeneity. This research direction is presented in [NLQO22]. The next chapter focuses on model heterogeneity as an inclusion policy for low-resource clients. It investigates the implications of client resource constraints on privacy given a reduced model complexity in low-resource clients. This work has been presented in [NLQO25]. The final area provides a solution to the data heterogeneity problem with distribution aware client selection. Applying this solution can mitigate spurious correlations and improve algorithmic fairness in FL. This research line has been described in [NFN+25]. [NLQO22] Németh, G. D., Lozano, M. A., Quadrianto, N., and Oliver, N. (2022). A Snapshot of the Frontiers of Client Selection in Federated Learning. Transactions on Machine Learning Research. [NLQO25] Németh, G. D., Lozano, M. A., Quadrianto, N., and Oliver, N. (2025). Privacy and Accuracy Implications of Model Complexity and Integration in Heterogeneous Federated Learning. IEEE Access, 13, 40258-40274. [NFN+25] Németh, G. D., Fani, E., Ng, Y. J., Caputo, B., Lozano, M. A., Oliver, N., and Quadrianto, N.(2025). FedDiverse: Tackling Data Heterogeneity in Federated Learning with Diversity-Driven Client Selection.
RUA. Repositorio Institucional de la Universidad de Alicante
Publication Otros (Other)

A Snapshot of the Frontiers of Client Selection in Federated Learning

  • Németh, Gergely Dániel
  • Lozano, Miguel Angel
  • Quadrianto, Novi
  • Oliver, Nuria
  • Lozano, Miguel Angel|||/items/413f111d-7a63-43f2-bb70-6421affe1647
Federated learning (FL) has been proposed as a privacy-preserving approach in distributed machine learning. A federated learning architecture consists of a central server and a number of clients that have access to private, potentially sensitive data. Clients are able to keep their data in their local machines and only share their locally trained model’s parameters with a central server that manages the collaborative learning process. FL has delivered promising results in real-life scenarios, such as healthcare, energy, and finance. However, when the number of participating clients is large, the overhead of managing the clients slows down the learning. Thus, client selection has been introduced as an approach to limit the number of communicating parties at every step of the process. Since the early naïve random selection of clients, several client selection methods have been proposed in the literature. Unfortunately, given that this is an emergent field, there is a lack of a taxonomy of client selection methods, making it hard to compare approaches. In this paper, we propose a taxonomy of client selection in Federated Learning that enables us to shed light on current progress in the field and identify potential areas of future research in this promising area of machine learning.