Instructor: Prof. O-Joun Lee

Office: Michael Hall T404

Office Hour:

e-mail: [email protected]

Website: https://nslab-cuk.github.io/

Lecture Time: Tuesday 2-3/Thursday 3

TA: Van Thuy Hoang

Lab: Sophie Barat Hall B348

Office Hour:

e-mail: [email protected]

TA: Sang Thanh Nguyen

Lab: Sophie Barat Hall B348

Office Hour:

e-mail: [email protected]

📜 Introduction

This course is designed to provide an in-depth exploration of Graph Neural Networks (GNNs), one of the most exciting areas of recent developments in machine learning and AI research. Students will gain a comprehensive understanding of GNNs, including their foundations, state-of-the-art models, and practical applications.

We will start with the basics of graph theory and Graph Convolutional Networks (GCNs), then move into more advanced topics, including Graph Attention Networks (GATs), structure-preserving GNNs, overcoming GCN limitations with models like GCNII, and Graph Transformer Networks. The course will also cover graph pooling techniques, graph autoencoders, and generative models for graphs. Moreover, we will explore handling dynamic graphs and heterogeneous graphs.

Throughout the course, we will place emphasis on the practical application of these models, focusing on various real-world scenarios and use cases. The course will mix theoretical lectures with practical exercises and coding assignments, giving students a balance between understanding the principles behind GNNs and applying them in practice.

By the end of the course, students will be able to:

• Understand the foundational concepts of graph theory and how they apply to machine learning.
• Understand the architecture, training, and use of various Graph Neural Network models, including GCNs, GATs, and Graph Transformers.
• Develop and train GNNs to solve real-world problems.
• Recognize the limitations of current GNNs and understand the solutions offered by state-of-the-art models.
• Analyze, evaluate, and compare different GNN models based on their performance, applicability, and limitations.
• Stay up-to-date with the latest research and developments in the field of GNNs.
• Apply graph neural networks to different fields, such as natural language processing, social network analysis, bioinformatics, and recommendation systems.

By the end of this course, students should be well-equipped to implement GNNs in their projects or research, as well as understand the current research landscape in the field of Graph Neural Networks.

☑️ Prerequisites

• 06837 Graph Mining

📚 Course Materials

References

🗓 Schedule