Fast Training of Sparse Graph Neural Networks on Dense Hardware.
Matej BalogBart van MerriënboerSubhodeep MoitraYujia LiDaniel TarlowPublished in: CoRR (2019)
Keyphrases
- neural network
- training process
- training algorithm
- feedforward neural networks
- real time
- densely connected
- weighted graph
- multi layer perceptron
- feed forward neural networks
- low cost
- neural network training
- pattern recognition
- image processing
- random walk
- graph representation
- neural network structure
- computer systems
- recurrent networks
- error back propagation
- artificial neural networks
- dense stereo
- gaussian graphical models
- directed acyclic graph
- bipartite graph
- graph theory
- feed forward
- multi layer
- self organizing maps
- structured data
- back propagation
- training set
- training examples
- training samples
- supervised learning
- motion field estimation
- genetic algorithm
- quasi cliques
- high dimensional
- fuzzy logic
- radial basis function network
- compressive sensing
- sparse representation
- neural network model