Towards Deepening Graph Neural Networks: A GNTK-based Optimization Perspective.
Wei HuangYayong LiWeitao DuRichard Y. D. XuJie YinLing ChenMiao ZhangPublished in: ICLR (2022)
Keyphrases
- neural network
- pattern recognition
- optimization process
- graph theory
- fuzzy logic
- optimization algorithm
- graph model
- graph structure
- highly non linear
- graph theoretic
- optimization problems
- random walk
- directed acyclic graph
- multi layer
- optimization methods
- neural network model
- global optimization
- directed graph
- combinatorial optimization
- graph representation
- genetic algorithm
- constrained optimization
- link analysis
- graph databases
- graph based algorithm
- min sum
- bipartite graph
- convex optimization
- graph matching
- feed forward
- self organizing maps
- back propagation
- sufficient conditions
- np hard
- multi objective
- viewpoint
- artificial neural networks