- Instructor: William L. Hamilton
- Term: Winter 2020
- When: Mondays and Wednesdays, 1:05-2:25pm
- Where: MC 12 Note the room change!
Graph representation learning (GRL) is a quickly growing subfield of machine learning that seeks to apply machine learning methods to graph-structured data. Machine learning on graphs is an important and ubiquitous task with applications ranging from drug design to friendship recommendation in social networks. This course will provide an introduction to graph representation learning, including matrix factorization-based methods, random-walk based algorithms, and graph neural networks. During the course, we will study both the theoretical motivations and practical applications of these methods.
A strong foundation in machine learning and proficiency with Python (or knowledge of multiple programming languages and the ability to quickly learn Python) is required. Knowledge of probabilities/statistics, calculus and linear algebra is required. Example courses at McGill providing sufficient background in probability are MATH-324. Machine learning background, as provided for example by COMP-551 or COMP-652 is required. If you have doubts regarding your background, please contact Will to discuss it.
List of topics (subject to minor changes)
- Basics of network and graph analysis
- Link prediction
- Graph and node classification
- Spectral clustering
- Node embeddings
- Knowledge graph embeddings
- Graph neural networks
- Graph signal processing
- Deep generative models of graphs
- Applications in computational biology and NLP