Federated learning (FL) enables collaborative model training across distributed Internet of Things (IoT) devices while preserving data privacy. However, in dynamic and open IoT environments, untrustworthy clients, adversarial attacks, and insufficient traceability hinder the robustness and security of FL systems. To ad...
Jiake Yin, Xianwei Zhou, Zhe Liu et al.· Journal of Cloud Computing A...· 0 citations
Audited reproducibility package for a five-seed comparison of FedAvg, FedProx,and SCAFFOLD under communication stress using UCI ElectricityLoadDiagrams20112014and Ausgrid Solar Home Electricity Data. The archive contains sanitized run-level outputs for 300 unique scientificexecutions, independently verifiable downstrea...
Manuel José Cabral dos Santos Reis· Zenodo (CERN European Organi...· 0 citations
Research artefact for the study FedStaP: Shared Feature Statistics and Prior-Calibrated Training for Non-IID Federated Intrusion Detection in IoT Networks. Contents. The full implementation (Hydra configurations, federated simulator, the proposed FedStaP method, ten global-model baselines, a personalised reference and...
Ahmed Fahad, Mohammed F. Alomari, Yazan Aljeroudi· Zenodo (CERN European Organi...· 0 citations
Abstract EN This document, produced with the assistance of ChatGPT o3 and GPT-5 Thinking, is released under the Apache 2.0 license. It is a voluntary defensive publication (prior art) and therefore enters the prior art upon release under the applicable patent statutes: EPC Art. 54(2) (European Patent Convention), Frenc...
Xavier Pillet· Zenodo (CERN European Organi...· 0 citations
Reach audiences
Advertise in front of researchers, engineers, and readers.
The segmentation of spectral domain optical coherence tomography (SD-OCT) images for diabetic macular edema (DME) using deep learning technology has key challenges such as data privacy, computational cost, and information uncertainty. To address these, we present an uncertainty-informed neural network within sequential...
Qiong Chen, Yanying Chen, Jinghui Wang et al.· Scientific Reports· 0 citations
Digital soil mapping has produced the first continent-wide baseline estimates of soil properties in Africa, yet existing soil information systems remain fragmented, poorly standardized, and largely confined to static offline spatial prediction. Their ability to capture near-real-time soil conditions or provide timely,...
Audited reproducibility package for a five-seed comparison of FedAvg, FedProx,and SCAFFOLD under communication stress using UCI ElectricityLoadDiagrams20112014and Ausgrid Solar Home Electricity Data. The archive contains sanitized run-level outputs for 300 unique scientificexecutions, independently verifiable downstrea...
Manuel José Cabral dos Santos Reis· Zenodo (CERN European Organi...· 0 citations
The rapid growth of IoT-enabled networks has intensified cyber threats, yet traditional intrusion detection systems (IDS) struggle to sustain high accuracy on high-dimensional traffic without heavy computational overhead. This paper addresses that dual challenge with an Adaptive Intrusion Detection System (AdaptIDS...
Rami Ahmad· Applied Computing and Inform...· 0 citations
"This dataset provides network traffic and physical sensor measurements intended to support research on cybersecurity, anomaly detection, machine learning, and multimodal security analysis in Cyber-Physical Systems (CPS). The network component contains 528,856 observations characterized by flow identifiers, source and...
A reproducible research framework for vertical federated learning on heterogeneous tabular data with explicit privacy boundaries, provenance and communication auditing.
Saurav Singla· Zenodo (CERN European Organi...· 0 citations
Each technology in the big data environment has its own strengths and weaknesses in terms of privacy protection; currently, no single technology can simultaneously meet all requirements, and there is a fundamental core conflict between privacy protection and data utility.
Zhen-Yuan Zhou· Applied and Computational En...· 0 citations
As data in scenarios such as medical, transportation and the Internet of Things continue to be dispersed to institutions and edge devices, how to balance the quality of model training and communication costs without concentrating raw data has become an important issue in the deployment of federated learning. Federated...
Chang Ma· Applied and Computational En...· 0 citations
Students in MIT’s Concourse program delve deeply into the human condition, debate challenging questions, and learn to develop judgment about issues that can’t be quantified.
Training AI agents with reinforcement learning can be challenging because their tools, context, and decision-making are managed by complex frameworks. Agent Lightning connects existing agents to RL training, making it easier to improve them without rebuilding them. The post Agent Lightning v1.0: A 3,500-Line Lightweight Agentic RL Framework for Training Agents with Real Harnesses appeared first on Microsoft Research.
MIT News · Artificial Intelligence· news.mit.eduOct 6, 2026
