Event BannerAbstract
In this lecture, we will delve into the world of Graph Neural Networks (GNNs). We will start by understanding the basics of graph theory and how it can be used to represent data. Then, we will introduce the concept of GNNs and how they can be used to learn from graph-structured data. We will cover different types of GNNs, including Graph Convolutional Networks (GCNs), Graph Attention Networks (GATs), and GraphSAGE. Finally, we will discuss some applications of GNNs in various domains, such as social network analysis, recommendation systems, and bioinformatics.
A three-hour lecture about Graph Neural Networks (GNNs) as part of the Delve Deep Learning Online course. The course is aimed at professionals, companies, and future entrepreneurs-researchers from different sectors of activity. The lecture will cover the basics of graph theory, the concept of GNNs, different types of GNNs, and applications of GNNs in various domains. The lecture will include both theoretical and practical aspects, with hands-on exercises to help participants understand the concepts better.
Outline
- Introduction
- Applications of networks
- Tasks you do with networks
- Defining a Graph
- Basic concepts of graph theory
- Community detection
- Importance detection
- Centrality or roles of nodes in the network
- Graph generation
- Random Walks
- First graph embeddings
- Based on factorization
- Based on random walks
- Graph Neural Networks
- Challenges
- Message Passing
- Computational Graphs
- Read-out layers
- Steps and notation
- Architectures
- GCN
- GraphSage
- GAT
- Hands-on
Adrián Arnaiz-Rodríguez
Artificial Intelligence PhD Student
ELLIS PhD Student in Algorithmic Fairness and Graph Neural Networks at ELLIS Alicante.