On the Unreasonable Effectiveness of Feature Propagation in Learning on Graphs With Missing Node Features.
Emanuele RossiHenry KenlayMaria I. GorinovaBenjamin Paul ChamberlainXiaowen DongMichael M. BronsteinPublished in: LoG (2022)
Keyphrases
- feature vectors
- learning algorithm
- image features
- selecting features
- learning process
- feature set
- learning tasks
- prior knowledge
- online learning
- supervised learning
- multiple features
- co occurrence
- predictive features
- feature relevance
- feature interactions
- topological features
- additional features
- feature values
- learning systems
- feature representation
- graph structure
- graph matching
- probabilistic model
- low level
- reinforcement learning
- feature extraction