Isometric Transformation Invariant and Equivariant Graph Convolutional Networks.
Masanobu HorieNaoki MoritaYu IharaNaoto MitsumePublished in: CoRR (2020)
Keyphrases
- average degree
- graph structure
- graph theory
- graph representation
- edge weights
- highly connected
- social networks
- directed graph
- small world
- structured data
- complex networks
- community discovery
- dynamic networks
- fully connected
- geometric transformations
- random walk
- network analysis
- weighted graph
- manifold learning
- rotation invariant
- random graphs
- image transformations
- transformation rules
- graph matching
- structural patterns
- sparse coding
- connected components