iSpLib: A Library for Accelerating Graph Neural Networks using Auto-tuned Sparse Operations.
Md Saidul Hoque AnikPranav BadheRohit GampaAriful AzadPublished in: WWW (Companion Volume) (2024)
Keyphrases
- neural network
- pattern recognition
- graph representation
- gaussian graphical models
- bipartite graph
- sparse data
- graph structure
- connected components
- random walk
- structured data
- fault diagnosis
- directed graph
- graph theoretic
- fuzzy systems
- graph theory
- graph search
- sufficient conditions
- graph data
- weighted graph
- adjacency matrix
- graph based algorithm
- directed acyclic
- artificial neural networks
- graph partitioning
- undirected graph
- spanning tree
- back propagation
- neural nets
- recurrent neural networks
- sparse coding
- feed forward
- sparse representation