Grafowe sieci neuronowe. Teoria i praktyka

Overview

This book offers a structured introduction to graph neural networks, combining the foundations needed to understand graph data with practical model implementation. It begins with graph-analysis tools in Python and selected topics from graph theory, then introduces graph neural networks and commonly used graph convolution layers.

The later chapters cover model design, data splits, training on larger graphs, and challenges such as oversmoothing. Practical examples show node, edge, and graph classification in different application settings.

Intended audience

The book is intended for data scientists, machine learning engineers, researchers, and students who know the basics of Python and machine learning and want to work with graph-structured data. The book is written in Polish.