Explaining compound activity predictions with a substructure-aware loss for graph neural networks.
Kenza AmaraRaquel Rodríguez-PérezJosé Jiménez-LunaPublished in: J. Cheminformatics (2023)
Keyphrases
- neural network
- graph data
- substructure discovery
- graph representation
- artificial neural networks
- graph theory
- random walk
- multi layer
- fuzzy logic
- structured data
- directed graph
- graph model
- genetic algorithm
- weighted graph
- recurrent neural networks
- graph structure
- graph matching
- feed forward
- fuzzy systems
- neural network model
- pattern recognition
- graph based algorithm
- multilayer perceptron
- regular expressions
- graph partitioning
- activation function
- control system
- graph structures
- frequent subgraph mining
- human activities