Machine Learning with Graphs (MLG)
This course is based on the course CS224W offered at the Stanford University. The course starts on Jan 10th, 2023 and lasts for ten weeks not including the breaks. The lectures will be recorded at Stanford on Tuesdays and Thursdays (≈three hours per week), and the videos will be available for download through this web page. Weekly consultations will be on Tuesdays evening (≈two hours per week), which will include short review of last two lectures by selected students and consultations on the coursework (Q&A). We will precisely follow the CS224W schedule, material, grading policy and other, but not the coursework (see below).
All course communication should be through Piazza! You should not contact or interact with Stanford in any way.
Students should check CS224W web site and this web page for updates frequently.
Complex data can be represented as a graph of relationships between objects. Such networks are a fundamental tool for modeling social, technological, and biological systems. This course focuses on the computational, algorithmic, and modeling challenges specific to the analysis of massive graphs. By means of studying the underlying graph structure and its features, students are introduced to machine learning techniques and data mining tools apt to reveal insights on a variety of networks. Tentative course outline is available on CS224W web page.
Students are expected to have the knowledge of basic computer science principles at a level sufficient to write a non-trivial computer program in Python, and are familiar with basic machine learning and data mining, probability theory and linear algebra.
There is no official text or book for the course. Course slides, handouts, readings and other material will be posted periodically on CS224W and this web page.
The books Graph Representation Learning by William L. Hamilton, Networks, Crowds, and Markets: Reasoning About a Highly Connected World by David Easley and Jon Kleinberg, and Network Science by Albert-László Barabási are recommended as optional reading.
Coursework & grading
Coursework consists of three homeworks, five programming assignments, and course project proposal, report and presentation. Students are encouraged to form groups of three or four for the course project.
Grading will be based 30% on homeworks (10% on each homework), 30% on programming assignments (6% on each assignment) and 40% on course project (10% on proposal, 25% on report and 5% on presentation). Extra credit can be assigned based on course participation and commitment.
Each student who submits at least three assignments will be graded at the end of the course.
All assignments will be available through CS224W web site and are usually due on Fridays at 9:00am (with some exceptions). Twice during the course you can take advantage of late periods. For an assignment due on Friday (Wednesday), late period expires on Tuesday (Sunday) at 9:00am.
All assignments must include a cover sheet with signed honor code and must be submitted through (1) Gradescope using course code XVNBVG and (2) eUcilnica using the options below. Failing to submit honor code or all developed programming code will result in 10% deduction.