Overview

This seminar introduces current research challenges at the intersection of machine learning and computer security. Students work in small teams on hands-on problems, gaining experience with research methods and learning how to identify and apply relevant ideas from the literature. Over the course of the semester, students tackle six challenges covering attacks and defenses against machine learning systems, as well as the use of learning-based techniques for practical computer security tasks.

Recommended Background

A strong background in machine learning and computer security is recommended. The course involves independent study, and students should be prepared to fill gaps in their prior knowledge as needed. Since much of the work is carried out remotely on a shared server, students should also have prior experience with, or be willing to independently learn, the necessary tools and workflows for remote development.

Organization

The course is organized around six challenges distributed over three waves. In each wave, students work in small teams for four to six weeks on two challenges: one focused on applying machine learning to security problems and one examining the security of machine learning systems. Challenges are run in a Kaggle-style competition format with a shared scoreboard, allowing teams to compare results throughout each wave. Students are given access to a remote server that provides the research problem and the necessary compute, and are expected to independently research relevant background, explore existing methods, experiment with different approaches, and develop and evaluate their own solutions.

Grading & Deliverables

Grades are based on the results achieved in the challenges and on the clarity with which results are documented and communicated. Challenges are solved in teams, with reports reflecting the team's work, while presentations are completed individually. There are two main deliverables:

• Technical reports. For each challenge, teams submit a short technical report (2 pages) written in the style of a mini research paper. The report briefly introduces the problem, describes the methodology, presents experimental results, and includes critical reflection. A report template will be provided and must be used. Technical reports and challenge results together account for 80% of the final grade.
• Presentations. Each student gives a 10-minute presentation on one selected challenge, based on the corresponding technical report. Presentations account for the remaining 20% of the final grade.