Physical Pooling Functions in Graph Neural Networks for Molecular Property Prediction.
Artur M. SchweidtmannJan G. RittigJana M. WeberMartin GroheManuel DahmenKai LeonhardAlexander MitsosPublished in: CoRR (2022)
Keyphrases
- neural network
- protein function prediction
- prediction model
- prediction accuracy
- graph structure
- random walk
- graph representation
- artificial neural networks to predict
- graph properties
- connected components
- fuzzy logic
- pattern recognition
- graph theory
- drug design
- global consistency
- graph model
- multi layer perceptron
- back propagation
- multilayer perceptron
- genetic algorithm
- three dimensional
- structured data
- spanning tree
- neural networks and support vector machines
- neural network ensemble
- graph theoretic
- prediction error
- artificial neural networks
- directed graph
- predictive model
- multi layer
- prediction algorithm
- weighted graph
- bipartite graph
- graph matching
- real world