Redundancy-Free Computation Graphs for Graph Neural Networks.
Zhihao JiaSina LinRex YingJiaxuan YouJure LeskovecAlex AikenPublished in: CoRR (2019)
Keyphrases
- neural network
- subgraph isomorphism
- graph matching
- graph representation
- directed graph
- graph structure
- graph theory
- graph databases
- graph construction
- graph model
- weighted graph
- pattern recognition
- graph structures
- graph theoretical
- labeled graphs
- graph theoretic
- graph mining
- structural pattern recognition
- graph properties
- adjacency matrix
- series parallel
- graph clustering
- graph isomorphism
- graph data
- graph partitioning
- graph classification
- graph search
- spanning tree
- bipartite graph
- graph kernels
- graph representations
- artificial neural networks
- random graphs
- undirected graph
- graph transformation
- dynamic graph
- evolving graphs
- reachability queries
- planar graphs
- community discovery
- graph patterns
- edge weights
- maximal cliques
- directed acyclic
- random walk
- finding the shortest path
- minimum spanning tree
- maximum independent set
- maximum common subgraph
- graph layout
- dense subgraphs
- connected graphs
- connected dominating set
- betweenness centrality
- web graph
- maximum clique
- query graph
- graph embedding
- real world graphs
- pattern mining
- small world
- relational structures
- median graph
- attributed graphs
- densely connected
- average degree
- frequent subgraphs
- proximity graph
- disk resident
- maximum cardinality
- polynomial time complexity
- directed acyclic graph
- back propagation
- shortest path