LEGNN trains noise-resistant GNNs by ensembling high- and low-probability predicted labels from randomly masked neighbor views, improving accuracy while cutting complexity relative to similarity-based reliable labeling.
Tax evasion detection with fbne-pu algorithm based on pncgcn and pu learning,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2024 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Training a Label-Noise-Resistant GNN with Reduced Complexity
LEGNN trains noise-resistant GNNs by ensembling high- and low-probability predicted labels from randomly masked neighbor views, improving accuracy while cutting complexity relative to similarity-based reliable labeling.