GOOD-D: On Unsupervised Graph Out-Of-Distribution Detection.
Yixin LiuKaize DingHuan LiuShirui PanPublished in: CoRR (2022)
Keyphrases
- false alarms
- object detection
- random walk
- detection method
- graph representation
- automatic detection
- detection algorithm
- directed graph
- probability distribution
- detection accuracy
- graph model
- supervised learning
- false positives
- power law
- bipartite graph
- graph theory
- weighted graph
- graph based algorithm
- random variables
- graph construction
- data sets
- spatial distribution
- semi supervised
- directed acyclic graph
- graphical models
- graph structure
- graph matching
- data driven
- structured data