ReGNN: a ReRAM-based heterogeneous architecture for general graph neural networks.
Cong LiuHaikun LiuHai JinXiaofei LiaoYu ZhangZhuohui DuanJiahong XuHuize LiPublished in: DAC (2022)
Keyphrases
- neural network
- special case
- management system
- pattern recognition
- fuzzy logic
- multi layer
- learning capabilities
- closely related
- directed acyclic graph
- network architecture
- graph structure
- radial basis function
- back propagation
- random walk
- real world
- heterogeneous networks
- graph representation
- heterogeneous environments
- graph model
- weighted graph
- graph theory
- neural nets
- software architecture
- image segmentation