Distributed Training of Large Graph Neural Networks with Variable Communication Rates.
Juan CerviñoMd Asadullah TurjaHesham MostafaNageen HimayatAlejandro RibeiroPublished in: CoRR (2024)
Keyphrases
- neural network
- communication overhead
- training process
- communication cost
- training algorithm
- feedforward neural networks
- spatially distributed
- computer networks
- distributed systems
- fully distributed
- pattern recognition
- multi layer perceptron
- cooperative
- multi party
- feed forward neural networks
- distributed computation
- peer to peer
- distributed control
- random walk
- artificial neural networks
- distributed network
- training set
- radial basis function network
- backpropagation algorithm
- training phase
- error back propagation
- graph structure
- fuzzy logic
- recurrent neural networks
- mobile agents
- feed forward
- communication systems
- supervised learning
- back propagation
- directed graph
- distributed environment
- graph representation
- communication networks
- graph mining
- training data
- multi agent
- genetic algorithm
- training samples
- training examples
- open systems
- exchange information
- information dissemination
- graph model
- multi layer