Graph Neural Network (GNN)
A neural network designed to operate on graph-structured data, where entities (nodes) are connected by relationships (edges).
How It Works
GNNs use message passing: each node aggregates information from its neighbors, updates its own representation, and repeats for multiple rounds. This captures the structural context around each node.
Applications
Social network analysis, drug discovery (molecular graphs), recommendation systems (user-item graphs), fraud detection (transaction graphs), traffic prediction, and knowledge graph reasoning.
Key Architectures
GCN (Graph Convolutional Network): Applies convolution-like operations on graphs. GAT (Graph Attention Network): Uses attention to weigh neighbor importance. GraphSAGE: Samples and aggregates neighbors for scalability.