6 April 2025 in Expert Systems with Applications
3 July 2024 in Expert Systems with Applications
28 November 2023 in Soft Computing
15 April 2025 in IEEE Transactions on Network and Service Management
Hierarchical federated learning (HFL) is a privacy-preserving distributed machine learning framework with a client-edge-cloud hierarchy, where multiple edge servers perform partial model aggregation to reduce costly communication with the cloud server. Nevertheless, most existing HFL methods require extensive iterative communication and public datasets, which not only increase communication overhead but also raise privacy and security concerns. Moreover, non-independent and identically distributed (non-IID) data among devices can significantly impact the accuracy of the global model in HFL. To address these challenges, we propose a multi-level one-shot HFL framework (MOHFL), which aims to improve the performance of the global model in a single communication round. Specifically, we employ conditional variational autoencoders (CVAEs) as local models and use the aggregated decoders to generate an IID training set for the global model, thereby mitigating the negative impact of non-IID data. We improve the performance of CVAEs under different levels of data heterogeneity through a dominant class-based data selection method. Subsequently, an edge aggregation scheme based on multi-teacher knowledge distillation and contrastive learning is proposed to aggregate the knowledge from local decoders to edge decoders. Extensive experiments on four real-world datasets demonstrate that MOHFL is very competitive against four state-of-the-art baselines under various settings.
11 December 2024 in Applied Intelligence
16 February 2026 in IEEE Internet of Things Journal
Federated learning (FL) is a distributed machine learning algorithm that enables multiple devices to collaboratively train a global model with data privacy protection. However, in real-world scenarios, the prevalent non-independent and identically distributed (non-IID) and long-tailed data across devices results in significant performance degradation in FL. Although numerous federated long-tailed learning (Fed-LT) methods have been proposed to address this issue, they typically rely on extensive iterative communication or public datasets, resulting in high communication overhead and potential privacy and security risks. To address these challenges, we propose a one-shot Fed-LT (OFLT) framework that aims to improve the global model’s performance in FL under non-IID and long-tailed data in a single communication round. Specifically, OFLT employs class center-based contrastive learning and feature decorrelation for discriminative representation learning on clients, enhancing the intraclass compactness and interclass separability. Subsequently, supervised distillation-based generator training and adaptive logit calibration are introduced on the server to mitigate class bias in the ensemble output of local models, and the calibrated knowledge is distilled into the global model. Extensive experiments on four image datasets demonstrate that OFLT outperforms state-of-the-art (SOTA) baselines under various settings, consistently improving not only overall accuracy but also performance on medium-shot and few-shot classes.