Descent Steps of a Relation-Aware Energy Produce Heterogeneous Graph Neural Networks.
Hongjoon AhnYongyi YangQuan GanDavid WipfTaesup MoonPublished in: CoRR (2022)
Keyphrases
- neural network
- pattern recognition
- graph structure
- artificial neural networks
- energy consumption
- graph theory
- graph theoretic
- back propagation
- graph model
- directed graph
- directed acyclic graph
- weighted graph
- recurrent neural networks
- graph matching
- feed forward
- self organizing maps
- random walk
- structured data
- bipartite graph
- graph representation
- spanning tree
- graph mining
- graph clustering
- low energy
- neural nets
- multilayer perceptron
- connected components
- sufficient conditions
- social networks
- genetic algorithm
- real world