Training Matters: Unlocking Potentials of Deeper Graph Convolutional Neural Networks.
Sitao LuanMingde ZhaoXiao-Wen ChangDoina PrecupPublished in: CoRR (2020)
Keyphrases
- convolutional neural networks
- graph representation
- convolutional network
- higher order
- training algorithm
- high order
- weighted graph
- graph theory
- graph structure
- training phase
- graph model
- structured data
- neural network
- connected components
- test set
- training samples
- random graphs
- training set
- graph matching
- training process
- graph mining
- np complete
- graph theoretic
- graph based algorithm
- graphical models