Skip to content
Log in

Similar papers

Conference Jul 2026

Improving Federated Learning under Non-IID Data through Clustered Ensembles and Adaptive Client Selection

Federated Learning (FL) enables collaborative model training without sharing raw data, but its effectiveness degrades under Non-IID client data and inefficient participation. In such settings, FedAvg may exhibit slow convergence, high variance across rounds, and elevated communication overhead. We propose a clustered ensemble framework to improve training stability and communication efficiency under label-skewed Non-IID distributions. Clients are clustered by data similarity, and one model is trained per cluster. At inference time, we employ a confidence-based ensemble that selects (or combines) cluster models, while adaptive client selection regulates participation within clusters to reduce redundant communication. Experiments on MNIST, Fashion-MNIST, CIFAR-10, and SVHN show more stable learning dynamics and lower communication cost while maintaining competitive accuracy compared to FedAvg. We also analyze how participation policies affect convergence behavior and client fairness.

Artur Sousa Freitas, A. T. Akabane, J. Estrella · 0 citations
2026

Asymmetric Partial Model Transmission for Federated Edge Learning

Federated learning (FL) applications normally employ large deep learning (DL) models, resulting in excessive communication overhead in the deployment of FL over resource-constraint mobile edge networks. To achieve better scalability for DL-based FL, we capitalize on both the asymmetric nature of mobile networks and the distinct effects of partial transmissions on FL training for the global and local models. We propose Fed-DynAmal, an FL framework that decreases the number of parameters transmitted in the uplink (clients-to-server) while concurrently achieving better model performance. The underlying idea is that each selected client sends a partial DL model to the server by omitting several sub-blocks from the trained local model. Crucially, we drop the assumption that transmitted local models can still be used for inference, thereby allowing for greater model variability. At the server, we introduce amalgamation, a process to merge different partial local models into an inference-viable full model. Essentially, amalgamation is a bridge for performing aggregation at the sub-block level. Interestingly, as the key takeaway, communication efficiency versus model performance is not necessarily a trade-off in FL: Our extensive experiments show that Fed-DynAmal can effectively improve communication efficiency while still concurrently achieving higher accuracy and enhanced robustness.

Zihan Chen, H. Yang, Tony Q. S. Quek et al. · 0 citations
Preprint Aug 2026

Beyond Parameter Space: NTK-Guided Personalized Aggregation for Robust Federated Learning

Local Inference Guided Aggregation for Heterogeneous Training Environments to Yield Enhancement Through Agreement and Regularization (LIGHTYEAR), a federated learning framework that performs update selection in function space using an NTK-based agreement score to characterize predictive behavior and determine a personalized aggregation set for each client.

Mirko Konstantin, S. Zachow, Anirban Mukhopadhyay · 0 citations
#machine learning Preprint Aug 2026

Efficient Resource Optimization for Split Federated Learning

This work establishes an efficient optimization framework for SFL under resource-constrained networks that jointly optimizes model splitting and resource allocation to minimize training cost, which is defined as the weighted sum of latency and energy costs.

Wei Wei, Xianhao Chen · 0 citations
Open access Jul 2026

Resilient federated learning under data and system heterogeneity via genetic algorithm-based group client selection

The process of selecting suitable clients to participate in Federated Learning (FL) remains a critical challenge due to non-IID heterogeneity in data distributions and heterogeneity in computational resources among participating clients. Existing client selection approaches typically assess clients individually using attributes such as model accuracy. Such individual-based client assessment ignores group effects, which are paramount for global model performance. In this paper, we propose a Genetic-Algorithm (GA) based client selection mechanism that is applicable to both horizontal FL (HFL) and vertical FL (VFL). Candidate client groups are evaluated by a multi-criteria fitness function that jointly models group data size, feature coverage, label coverage, class balance, feature-distribution divergence, feature importance, computational power, reputation, accuracy, and outlier proportion. The GA performs an efficient search over the combinatorial space of client subsets. The proposed approach is evaluated using image classification, diabetes prediction, and rain prediction tasks. The proposed approach improves global model accuracy and accelerates convergence. On the MNIST dataset, it peaks at 98.37%, outperforming DSCS (97.27%) and FedMCCS (97.07%). On the diabetes dataset, it achieves 88.37%, while DSCS and FedMCCS achieve 87.15% and 86.64%, respectively. On the rain prediction task, it attains a peak accuracy of 83.57% and converges to 83.36%, compared with 81.98% for FedMCCS and 80.38% for DSCS.

Sani Umar, Ahmed Alagha, R. Mizouni et al. · 0 citations
2026

Toward Efficient Semi-Asynchronous Federated Learning: A Multi-Factor Grouping and Dual-Level Selection Scheme Under Heterogeneous Environments

Semi-Asynchronous Federated Learning (SAFL) takes advantage of both synchronous and asynchronous FLs to train models. However, existing works in semi-asynchronous FLs fail to fully account for heterogeneities in both data and devices. To address these issues, we first propose a Clustered SAFL (CSAFL) framework and theoretically analyze its convergence loss. Then, a convergence loss minimization problem is formulated under the considerations of heterogeneities in device resources, fairness of cluster selection, and data heterogeneity. To address this complex problem due to nonlinearities and multi-dimensional decision variables, we first design a device clustering algorithm based on both devices’ model parameter differences and gradient directions between local and global models. Then, the original loss minimization problem is transformed into inter- and intra-cluster selection problems. For the inter-cluster selection problem, to ensure fairness, we propose a reinforcement learning-driven Lyapunov approach to fairly select clusters, where reinforcement learning (RL) is used to supervise the cluster selection results by the Lyapunov method. For the intra-cluster selection, we convert it into a constrained multi-armed bandit (MAB) problem in order to let devices within a cluster submit models synchronously. Then, a two-stage Upper Confidence Bound (UCB) scheme is proposed to obtain device selection results. Extensive numerical results with baselines show that our approach achieves up to 25% higher accuracy.

Gang Li, Yuhang Chen, Jun Cai et al. · 0 citations