Get Started
Home
Topics
Search
Library
Research questionHow should federated aggregation weight hospitals when local data size does not reflect diagnostic informativeness?Hospitals may contribute very different numbers of radiographs, but sample count alone does not indicate how informative each local dataset is for a shared diagnostic model. This can allow larger institutions to exert disproportionate influence during training.
AI
Computer Vision
Health
Machine Learning
Statistical Machine Learning
Latest papersRecent research connected to this question, newest first.FedDRAW: Federated Dual Reputation Annealing Weighting for Heterogeneous Multi-Institutional Chest Radiograph ClassificationThe source concerns federated chest-radiograph classification with patient data kept locally and aggregation performed across communication rounds. Evidence comes from 12 simulated client-partition scenarios using CheXpert and ChestMNIST, comparing eight aggregation methods under identical local-training settings with AUC and geometric mean sensitivity-specificity; it does not directly establish performance in real hospital deployment.research paper · Sep 4, 2026
Related questions
How can differentially private federated learning preserve faithful explanations without sacrificing predictive utility?How can pediatric chest X-ray classifiers preserve usable decisions across countries when calibration and source thresholds shift?How can federated learning protect client updates from inference while resisting Byzantine manipulation?How can federated learning personalize regional query understanding while preserving privacy and avoiding transformer training failures?