MixHop: Higher-Order Graph Convolutional Architectures via Sparsified Neighborhood Mixing.
Sami Abu-El-HaijaBryan PerozziAmol KapoorHrayr HarutyunyanNazanin AlipourfardKristina LermanGreg Ver SteegAram GalstyanPublished in: CoRR (2019)
Keyphrases
- higher order
- neighborhood graph
- high order
- natural images
- neighborhood structure
- higher order relations
- graph structure
- graph representation
- affinity graph
- graph theory
- random walk
- graph based algorithm
- minimal spanning tree
- connected components
- unsupervised feature learning
- low order
- graph model
- lower order
- pairwise
- markov random field
- structured data
- graph matching
- directed graph
- neighborhood information
- graph construction
- sparse coding
- conditional random fields
- neural network
- directed acyclic graph
- knn
- stable set
- manifold learning
- deep learning
- graph clustering
- graph data
- weighted graph
- spanning tree