GINNs: Graph-Informed Neural Networks for Multiscale Physics.
Eric Joseph HallSøren TaverniersMarkos A. KatsoulakisDaniel M. TartakovskyPublished in: CoRR (2020)
Keyphrases
- multiscale
- neural network
- graph theory
- pattern recognition
- genetic algorithm
- graph representation
- image representation
- coarse to fine
- directed graph
- dependency graph
- artificial neural networks
- random walk
- activation function
- graph structure
- feed forward
- image processing
- wavelet transform
- computer science
- graph databases
- undirected graph
- graph mining
- fuzzy systems
- multi layer
- multiscale analysis
- multilayer perceptron
- connected components
- fault diagnosis
- scale space
- machine learning
- bipartite graph
- neural nets
- recurrent neural networks
- directed acyclic graph
- filter bank
- expert systems