—
GraphSAGE
by Stanford University
An inductive representation learning framework for large graphs that samples and aggregates node features from local neighborhoods, enabling embeddings for unseen nodes.
Graph and Network Models
Operate on graph-structured data, learning from nodes, edges, and their relationships.
by Stanford University
An inductive representation learning framework for large graphs that samples and aggregates node features from local neighborhoods, enabling embeddings for unseen nodes.
by Mila / University of Cambridge
A neural network architecture using masked self-attention on graph-structured data, learning different weights for different neighbor nodes.
by Amazon Web Services
A framework for deep learning on graphs supporting PyTorch and TensorFlow, with optimized message passing primitives and distributed training.
by University of Amsterdam
A neural network architecture for semi-supervised learning on graph-structured data using efficient spectral approximations.