Deep Lagrangian Constraint-based Propagation in Graph Neural Networks.
Matteo TiezziGiuseppe MarraStefano MelacciMarco MagginiPublished in: CoRR (2020)
Keyphrases
- neural network
- graph theory
- pattern recognition
- graph based algorithm
- graph structure
- graph model
- graph theoretic
- graph representation
- constraint satisfaction
- connected components
- multilayer perceptron
- fuzzy logic
- structured data
- feed forward
- artificial neural networks
- random graphs
- graph data
- graph partitioning
- global consistency
- recurrent neural networks
- multi layer
- fault diagnosis
- fuzzy systems
- graph mining
- bipartite graph
- directed graph
- numerical methods
- back propagation
- sufficient conditions
- linear programming
- optimal solution
- learning algorithm