DistGNN: scalable distributed training for large-scale graph neural networks.
Vasimuddin MdSanchit MisraGuixiang MaRamanarayan MohantyEvangelos GeorganasAlexander HeineckeDhiraj D. KalamkarNesreen K. AhmedSasikanth AvanchaPublished in: SC (2021)
Keyphrases
- scalable distributed
- neural network
- training process
- training algorithm
- neural network training
- feedforward neural networks
- pattern recognition
- multi layer perceptron
- recurrent networks
- feed forward neural networks
- graph matching
- directed graph
- back propagation
- neural network structure
- graph representation
- error back propagation
- weighted graph
- backpropagation algorithm
- graph structure
- genetic algorithm
- supervised learning
- graph theory
- graph based algorithm
- training set
- artificial neural networks
- fuzzy logic
- random walk
- test set
- structured data
- multi layer
- recurrent neural networks
- spanning tree
- training phase
- training samples
- graph model
- fault diagnosis
- classifier training
- web pages
- neural nets
- real world
- data sets