Isometric Transformation Invariant and Equivariant Graph Convolutional Networks.
Masanobu HorieNaoki MoritaToshiaki HishinumaYu IharaNaoto MitsumePublished in: ICLR (2021)
Keyphrases
- average degree
- dynamic networks
- graph structure
- fully connected
- small world
- graph model
- euclidean space
- random walk
- graph representation
- weighted graph
- directed graph
- graph matching
- community detection
- rotation invariant
- community structure
- edge weights
- connected components
- image transformations
- structural patterns
- graph structures
- biological networks
- overlapping communities
- complex networks
- graph layout
- graph theory
- dimensionality reduction
- network structure
- deep learning
- graph data
- graphical representation
- sparse coding