Batched Sparse Matrix Multiplication for Accelerating Graph Convolutional Networks.
Yusuke NagasakaAkira NukadaRyosuke KojimaSatoshi MatsuokaPublished in: CoRR (2019)
Keyphrases
- matrix multiplication
- sparse coding
- densely connected
- graph theory
- community discovery
- message passing
- high dimensional
- sparse representation
- graph structures
- small world
- gaussian graphical models
- social networks
- overlapping communities
- weighted graph
- network structure
- edge weights
- directed graph
- deep learning
- complex networks
- fully connected
- bipartite graph
- power law
- distributed memory
- higher order
- software engineering
- dynamic programming
- image processing