Distilling Influences to Mitigate Prediction Churn in Graph Neural Networks.
Andreas RothThomas LiebigPublished in: ACML (2023)
Keyphrases
- neural network
- prediction accuracy
- artificial neural networks to predict
- random walk
- prediction model
- directed graph
- pattern recognition
- artificial neural networks
- neural networks and support vector machines
- neural network ensemble
- graph theory
- directed acyclic graph
- recurrent neural networks
- prediction algorithm
- predictive model
- multi layer perceptron
- graph structure
- prediction error
- graph model
- neural network model
- spanning tree
- weighted graph
- graph matching
- fuzzy systems
- graph partitioning
- bipartite graph
- neural nets
- structured data
- protein function prediction
- multilayer perceptron
- connected components