Accelerating SpMM Kernel with Cache-First Edge Sampling for Graph Neural Networks.
Chien-Yu LinLiang LuoLuis CezePublished in: CoRR (2021)
Keyphrases
- neural network
- weighted graph
- undirected graph
- edge weights
- disjoint paths
- graph kernels
- pattern recognition
- graph theory
- neural network model
- graph representation
- random sampling
- graph structure
- support vector
- query processing
- sample size
- feature space
- vertex set
- graph model
- kernel function
- database systems
- back propagation
- training data
- spanning tree
- artificial neural networks
- shortest path
- hit rate
- edge detector
- data access
- directed graph
- connected components
- monte carlo
- graph embedding
- graph laplacian
- beltrami flow
- sampling methods
- graph partitioning
- edge information
- prefetching
- graph matching
- similarity function
- fuzzy logic
- high dimensional