Exploring the Use of Dataflow Architectures for Graph Neural Network Workloads.
Ryien HosseiniFilippo SiminiVenkatram VishwanathRamakrishnan SivakumarSanjif ShanmugaveluZhengyu ChenLev ZlotnikMingran WangPhilip ColangeloAndrew DengPhilip LassenShukur PathanPublished in: ISC Workshops (2023)
Keyphrases
- neural network
- graph structure
- random walk
- database systems
- graph transformation
- pattern recognition
- graph theory
- artificial neural networks
- directed graph
- data flow
- bp neural network
- bipartite graph
- neural network is trained
- fuzzy artmap
- graph representation
- parallel computing
- back propagation
- graph model
- network architecture
- neural network model
- self organizing maps
- graph theoretic
- structured data
- graph based algorithm
- pairwise
- weighted graph
- directed acyclic graph
- graph matching
- fuzzy neural network
- spanning tree
- access patterns
- activation function