A Theoretical Investigation of Graph Degree as an Unsupervised Normality Measure.
Çaglar AytekinFrancesco CricriLixin FanEmre AksuPublished in: CoRR (2018)
Keyphrases
- theoretical investigation
- graph representation
- distance measure
- structured data
- probability measure
- data driven
- graph theory
- pairwise
- semi supervised
- graph connectivity
- unsupervised learning
- degree distribution
- graph model
- pointwise mutual information
- data sets
- graph based algorithm
- link analysis
- graph structure
- graph matching
- graphical models
- supervised learning
- similarity measure