Graph Neural Networks Use Graphs When They Shouldn't.
Maya Bechler-SpeicherIdo AmosRan Gilad-BachrachAmir GlobersonPublished in: CoRR (2023)
Keyphrases
- neural network
- graph theory
- graph representation
- graph structure
- weighted graph
- graph matching
- labeled graphs
- graph theoretic
- graph clustering
- graph databases
- graph construction
- graph classification
- graph partitioning
- graph theoretical
- graph structures
- directed graph
- adjacency matrix
- graph mining
- series parallel
- graph model
- graph search
- graph properties
- structural pattern recognition
- random graphs
- bipartite graph
- graph representations
- pattern recognition
- reachability queries
- spanning tree
- graph isomorphism
- graph data
- undirected graph
- graph transformation
- subgraph isomorphism
- finding the shortest path
- disk resident
- adjacency graph
- web graph
- connected graphs
- graph kernels
- maximum common subgraph
- minimum spanning tree
- maximum clique
- inexact graph matching
- dynamic graph
- edge weights
- artificial neural networks
- graph embedding
- directed acyclic
- planar graphs
- evolving graphs
- topological information
- dense subgraphs
- maximal cliques
- graph layout
- connected dominating set
- graph drawing
- community discovery
- fully connected
- social graphs
- strongly connected
- graph patterns
- maximum independent set
- random walk
- densely connected
- degree distribution
- maximum cardinality
- neighborhood graph
- complex networks
- structured data