DP-DGAD pretrains on labeled dynamic graphs and adapts to unlabeled ones through evolving normal/abnormal prototypes and pseudo-labels, reporting large AUROC/AUPRC gains on eight target datasets.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
REJECT 1representative citing papers
citing papers explorer
-
DP-DGAD: A Generalist Dynamic Graph Anomaly Detector with Dynamic Prototypes
DP-DGAD pretrains on labeled dynamic graphs and adapts to unlabeled ones through evolving normal/abnormal prototypes and pseudo-labels, reporting large AUROC/AUPRC gains on eight target datasets.