Descent Steps of a Relation-Aware Energy Produce Heterogeneous Graph Neural Networks.
Hongjoon AhnYongyi YangQuan GanTaesup MoonDavid P. WipfPublished in: NeurIPS (2022)
Keyphrases
- neural network
- graph representation
- pattern recognition
- artificial neural networks
- random walk
- recurrent neural networks
- directed graph
- fuzzy logic
- graph model
- directed acyclic graph
- graph theory
- multilayer perceptron
- neural network model
- dependency graph
- graph partitioning
- graph matching
- genetic algorithm
- energy consumption
- back propagation
- fuzzy systems
- energy saving
- graph theoretic
- energy minimization
- weighted graph
- graph structure
- structured data
- fault diagnosis
- multi layer
- bipartite graph
- graph cuts
- sensor networks
- image segmentation
- neural nets
- graph structures
- minimum energy
- global exponential stability
- feed forward