FedTGNN-SS delivers strong AUROC on GDM and related datasets even at 80% label missingness per site by combining local k-NN graphs, adaptive GNNs, prototype pseudo-labeling, and privacy-safe centroid sharing.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Federated Semi-Supervised Graph Neural Networks with Prototype-Guided Pseudo-Labeling for Privacy-Preserving Gestational Diabetes Mellitus Prediction
FedTGNN-SS delivers strong AUROC on GDM and related datasets even at 80% label missingness per site by combining local k-NN graphs, adaptive GNNs, prototype pseudo-labeling, and privacy-safe centroid sharing.