54369-01, Spring 2023

Instructor: Prof. O-Joun Lee

Office: Michael Hall T404

Office Hour:

e-mail: [email protected]

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

Lecture Time: Tuesday 7-8/Thursday 7

📜 Introduction

Graph neural networks, or GNNs, are a class of machine learning models that are designed to work with data that is represented in the form of graphs. They have become increasingly popular in recent years due to their ability to effectively process and analyze graph-structured data.

In this lecture, we will start by introducing the fundamental concepts of GNNs, including the mathematical foundations of graph theory and graph convolution. We will then move on to review state-of-the-art GNN models and their key contributions. Finally, we will discuss the pros and cons of GNNs. This lecture will be interactive, and I encourage you to ask questions and engage in the discussion throughout the class.

1. Introducing fundamental concepts of graph neural networks
   • Overview of graph neural networks
   • Mathematical foundations of graph theory and graph convolution
   • Explanation of the key components of GNNs, including node embeddings, graph convolution layers, and graph pooling
2. Reviewing papers for state-of-the-art graph neural network models
   • In-depth analysis of recent research papers on GNNs
   • Discussion of the key contributions of each paper
   • Comparison of different GNN models in terms of their performance and applicability
3. Discussing pros and cons of the graph neural network models
   • Comparison of GNNs with other machine learning techniques for graph data
   • Discussion of the limitations and challenges of GNNs

📚 Course Materials

References

🗓 Schedule

Weekly Schedule

🏆 Evaluation

Evaluation Criteria

Attendence: 10% Assignments: 20% Mid-term Exam: 35% Final Exam: 35%

Tentative Grading Scale

A 80~100 B 60~80 C ~60 F Absence from exams

Assignment Submission

Please make sure to submit all assignments through the online learning platform before the designated deadline. Grading for assignments will be based on a scale of A (10 points), B (8 points), C (6 points), and F (0 points) for any assignments that are not submitted on time.

Late Assignment Submission Policy

Late submissions will be accepted, but the maximum grade that can be achieved for any assignments submitted after the deadline is B (8 points).

😢 Plagiarism

Please be aware that plagiarism will not be tolerated in this class. All assignments and exams must be completed and submitted by the individual student, and any plagiarism will be dealt with accordingly. Be sure to properly cite any sources used in your work and avoid any form of academic dishonesty.