This course provides a technical introduction to Graph Neural Networks (GNNs), a class of models designed to perform inference on data structured as graphs. You will learn the principles behind how GNNs operate, from message passing mechanisms to specific architectures like Graph Convolutional Networks (GCNs) and Graph Attention Networks (GATs). The material covers the full process of building a GNN model: representing graph data, defining the network architecture, training the model, and evaluating its performance on common graph-related tasks. Practical sections focus on implementation using the PyTorch Geometric library.
Prerequisites Python & ML basics
Level:
Graph Data Representation
Represent graph-structured data for machine learning applications using adjacency and feature matrices.
GNN Mechanics
Explain the core message passing mechanism that forms the basis of most GNN architectures.
GNN Architectures
Differentiate and implement foundational GNN models, including GCNs, GraphSAGE, and GATs.
Model Training
Construct a complete training and evaluation pipeline for a GNN on tasks like node classification.
Practical Implementation
Build and train GNNs efficiently using the PyTorch Geometric library.
© 2026 ApX Machine LearningEngineered with