SkipNode: On Alleviating Over-smoothing for Deep Graph Convolutional Networks.
Weigang LuYibing ZhanZiyu GuanLiu LiuBaosheng YuWei ZhaoYaming YangDacheng TaoPublished in: CoRR (2021)
Keyphrases
- deep learning
- graph structures
- average degree
- fully connected
- dynamic networks
- graph model
- network analysis
- graph representation
- highly connected
- social graphs
- directed graph
- unsupervised feature learning
- graph theory
- bipartite graph
- graph layout
- random walk
- social networks
- biological networks
- small world
- edge weights
- graph theoretic
- graph mining
- weighted graph
- connected components
- complex networks
- graphical representation
- power law
- graph data
- neural network
- minimum spanning tree
- graph databases
- graph structure
- community discovery
- network structure
- structured data
- learning algorithm