Abstract What happens when algebra usurps political power? Studies in digital mobility predominantly focus on utilitarian geometric route optimization. However, this paradigm overlooks the fact that navigation algorithms operate as concealed normative actors, thereby distributing the right to the city. Drawing upon a s...
Oleksandr Hlushenko· Zenodo (CERN European Organi...· 0 citations
A novel Federated Anomaly Detection for Heterogeneous Smart Home Environment-based Zero-Trust Micro-Segmentation (CloudShield-IoT) framework is introduced, demonstrating strong resilience against Byzantine adversarial attacks.
Cross-border healthcare data sharing is critical to advance precision treatment, rare illness research, and global public health surveillance. However, legislative limits, privacy concerns, and institutional mistrust severely impede centralized data pooling across nations. Federated learning (FL) has emerged as a poten...
Reference implementation of SWIFT, a wait-free decentralized federated learning algorithm in which clients never block on one another: each broadcasts its model to its neighbours and averages in whatever has arrived. SWIFT is the first asynchronous decentralized federated learning algorithm to obtain the optimal O(1/sq...
Marco Bornstein, Tahseen Rabbani, Evan Wang et al.· Zenodo (CERN European Organi...· 0 citations
Reach audiences
Advertise in front of researchers, engineers, and readers.
Healthcare information technology (IT) infrastructure has grown substantially in complexity alongside the accelerating adoption of electronic health records, clinical decision support systems, telehealth platforms, and patient-facing digital services. Managing this infrastructure reliably under high availability expect...
Nayan Kumar Sureshbhai Patel· Journal of Information Syste...· 0 citations
The majority of the work now in publication discusses safeguard healthcare data however, the data modality—text, photos, audio, and sound—was not particularly addressed. Additionally, it was unclear how to safeguard while taking into account the differences in feature extraction depending on the input modalities. To de...
ReSIDS / ereno-Adaptativo is an intrusion-detection pipeline that unifies centralized, federated and gossip learning in a single run with round-by-round mode switching, built on Flower, scikit-learn and XGBoost. Version 2 adds: a distributed, automatic FL↔GL switch driven by decentralized failure detection (fail-fast F...
CassioBatista, Paulo Coelho, Silvio E. Quincozes et al.· Zenodo (CERN European Organi...· 0 citations
Abstract What happens when algebra usurps political power? Studies in digital mobility predominantly focus on utilitarian geometric route optimization. However, this paradigm overlooks the fact that navigation algorithms operate as concealed normative actors, thereby distributing the right to the city. Drawing upon a s...
Oleksandr Hlushenko· Zenodo (CERN European Organi...· 0 citations
Results show that clipping, perturbation, and the learning procedure jointly influence the observed privacy–utility behavior of centralized FL and sequential Decentralized Federated Learning.
AlsharifHasan Mohamad Aburbeian, M. Fernández-Veiga, A. Fernández-Vilas et al.· Future Internet· 0 citations
Transfer learning via fine-tuning pre-trained transformer models has gained significant suc- cess in delivering state-of-the-art results across various NLP tasks. In the absence of central- ized data, Federated Learning (FL) can benefit from distributed and private data of the FL edge clients for fine-tuning. However,...
Phishing websites are continuously modified to imitate legitimate services, evade static blacklists, and exploit user trust. Machine-learning-based detection can identify previously unseen patterns, but conventional centralized training requires sensitive URL, browsing, webpage, and behavioral information to be collect...
Sonal Kotkar, Siddhi Pawar, Tanvi Khadse et al.· International journal of inn...· 0 citations
StudentSim is a research system for evaluating and improving AI tutoring tools before they are tested with learners. It models how individuals may engage with tutors, giving feedback to improve tutor performance across subjects like chess, writing, and mathematics. The post Simulating student learning for better AI tutoring appeared first on Microsoft Research.
MIT News · Artificial Intelligence· news.mit.eduOct 7, 2026
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
