Format:
- There are no exams.
- The course is set up as 5 modules
- Introduction (2 classes)
- Intro to Privacy (~5 classes)
- Intro to Fairness (~5 classes)
- Algorithms for Privacy (6 classes)
- Algorithms for Fairness (6 classes)
- Modules 1-3: Instructors will lead the discussion with lectures
- Modules 4-5: Students read a paper and lead the discussion in the class.
Graded Student Work:
- Modules 2 and 3:
- Each student will independently work on a mini-project/homework assignment.
- Modules 4-5:
- Each student will individually submit a mini-critique for 10 papers that will be discussed in class. A mini-critique is like a peer-review at a conference and the very least should have:
- Summary: Motivation + Problem + Approach + Result
- 3 Strengths
- 3 Weaknesses
- 1-2 students will be assigned to lead the discussion of each paper.
- Each student will individually submit a mini-critique for 10 papers that will be discussed in class. A mini-critique is like a peer-review at a conference and the very least should have:
- Students will participate in an individual or group research project. Projects can focus on developing new theory/algorithms for privacy/fairness, or on implementing/adapting known algorithms to a real application setting.
- More information about project ideas is forthcoming.
Grading:
- Homework/mini-projects: 20 each
- Mini-critiques: 10
- Class participation: 10
- Project: 40
Standards of Conduct:
Under the Duke Community Standard, you are expected to do your own work–individually for homework/mini-projects, and with your team for the project. We will use the whiteboard policy for homework/mini-projects — you are allowed to discuss solutions with your peers, but you must write submissions/code independently (and indicate in your submission any assistance you received). Any assistance received that is not given proper citation will be considered a violation of the Standard.