Machine Learning with Graphs

Course Overview

This course explores machine learning on graph data. While Graph Neural Networks (GNNs) are a core focus, we also cover other neural network architectures such as CNNs and Transformers. The course is designed to help students connect different areas of machine learning, understanding both theory and practice.

Learning Approach

• Emphasis on applications and innovation, not just theory.
• Students learn to leverage data features for representation learning and adapt model architectures to suit different tasks and heuristics.
• Covers foundations, practical applications, and extensions of basic principles.

Structure & Requirements

• No exams.
• Attendance is expected.
• Students complete open-ended assignments and a course project.
• Prerequisites: Python programming. Prior machine learning knowledge is helpful but not required.

Goals

By the end of the course, students will have:

• A strong understanding of graph-based and general neural network architectures.
• Hands-on experience in applying and modifying models for different tasks.
• The ability to reason about data and model design in a flexible, innovative way.

Tentative Schedule

Schedule may change depending on class progress.