Comparison of Atom Representations in Graph Neural Networks for Molecular Property Prediction.
Agnieszka PochaTomasz DanelSabina PodlewskaJacek TaborLukasz MaziarkaPublished in: IJCNN (2021)
Keyphrases
- neural network
- protein function prediction
- prediction accuracy
- artificial neural networks to predict
- van der waals
- directed graph
- multi layer perceptron
- graph theory
- prediction model
- symbolic knowledge
- artificial neural networks
- pattern recognition
- fault diagnosis
- random walk
- graph properties
- fuzzy logic
- vector representation
- prediction algorithm
- three dimensional
- drug design
- protein structure prediction
- graph representation
- structured data
- predictive model
- graph theoretic
- spanning tree
- training process
- multi layer
- graph representations
- weighted graph
- distributed representations
- bipartite graph
- neural network ensemble
- graph matching
- microarray