PrivMaskFL: A private masking approach for heterogeneous federated learning in IoT

Article Computer Science, Information Systems

Real-time trajectory privacy protection based on improved differential privacy method and deep learning model

Jing Xiong et al.

Summary: Accurate and real-time trajectory data publishing is important for providing users with up-to-date information for travel planning. Improper publishing of location information can threaten users' privacy. This study proposes a data publishing method based on deep learning and differential privacy models to secure spatiotemporal trajectory data. Experimental results show that the proposed method achieves better results than traditional methods while satisfying differential privacy.

JOURNAL OF CLOUD COMPUTING-ADVANCES SYSTEMS AND APPLICATIONS (2022)

Add to Collection

Article Computer Science, Information Systems

FedMCCS: Multicriteria Client Selection Model for Optimal IoT Federated Learning

Sawsan AbdulRahman et al.

Summary: This article introduces FedMCCS, a multi-criteria client selection method for addressing the issue of limited client resources leading to decreased model accuracy in federated learning. FedMCCS outperforms other approaches in reducing communication rounds, maximizing the number of clients, handling the least number of discarded rounds, and optimizing network traffic.

IEEE INTERNET OF THINGS JOURNAL (2021)

Add to Collection

Review Radiology, Nuclear Medicine & Medical Imaging

Federated learning: a collaborative effort to achieve better medical imaging models for individual sites that have small labelled datasets

Dianwen Ng et al.

Summary: Federated learning is a new learning framework that allows different medical institutions to collaboratively train a global model to improve the application of AI in radiology. By aggregating training results from multiple sites, the model receives better supervision and global detection capabilities, while maintaining patient privacy.

QUANTITATIVE IMAGING IN MEDICINE AND SURGERY (2021)

Add to Collection

Article Computer Science, Information Systems

Client Selection for Federated Learning With Non-IID Data in Mobile Edge Computing

Wenyu Zhang et al.

Summary: The paper introduces a novel FL algorithm called CSFedAvg to alleviate the accuracy degradation caused by non-IID data, by utilizing weight divergence to recognize the non-IID degrees of clients and selecting clients with lower non-IID data frequency to train models more frequently. Simulations using publicly-available datasets show that the proposed FL algorithm improves training performance compared to existing protocols.

IEEE ACCESS (2021)

Add to Collection

Article Computer Science, Information Systems