Comparison of Atom Representations in Graph Neural Networks for Molecular Property Prediction.
Agnieszka PochaTomasz DanelLukasz MaziarkaPublished in: CoRR (2020)
Keyphrases
- neural network
- protein function prediction
- prediction accuracy
- graph properties
- van der waals
- pattern recognition
- artificial neural networks to predict
- prediction model
- prediction error
- graph structure
- neural network ensemble
- graph theory
- fuzzy logic
- random walk
- feed forward
- directed graph
- symbolic knowledge
- neural nets
- bipartite graph
- artificial neural networks
- structured data
- genetic algorithm
- three dimensional
- chaotic time series
- protein structure prediction
- graph theoretic
- graph representation
- multi layer perceptron
- graph partitioning
- graph databases
- graph model
- self organizing maps
- connected components
- multilayer perceptron
- graph matching
- spanning tree
- graphical representation
- prediction algorithm
- drug design
- dna computing
- multi layer
- fault diagnosis
- recurrent neural networks