Time : Mondays and Fridays. 1:30pm - 3:00pm
Location : ICCS 246
Instructor : Mi Jung Park
Email : mijungp@cs.ubc.ca (use [CS532P Privacy in ML] in subject line)
Zoom: https://ubc.zoom.us/j/9920729429?pwd=VWVkZ0ZFd1o5OTJkd2dKN3dpWXhlQT09
The course is generally a seminar, in which each student gets to present one paper per term (See Presentation for details). The student's presentation will be peer-to-peer graded by other students in the class. The instructor will also teach a few times for general introductions of certain subtopics, e.g., when introducing the basics of differential privacy, the methods for interpretability/explainability, and the notions of algorithmic fairness. There are no homework assignments and exams. However, you are required to complete a course project, which will be a significant amount of work to do during the term (See Project for details).
The relative weighting of the individual contributors to the grade is as follows:
Participation: 10%
Paper Presentation: 30% (Peer-to-peer grading)
Project: 60%
of which
Proposal: 20%
Report: 70%
Final presentation: 10% (Peer-to-peer grading)
I will be generous with grading, as long as you do your part diligently (i.e., presentation and course project).
Graduate level of "Probability and Random Processes"
Undergrad level of Linear Algebra
Machine Learning (Similar to the level taught in CPSC 340 and/or CPSC 440/540)
How to write code in Python and PyTorch
If you don't have these pre-requisites fulfilled, probably it's hard to follow what we discuss during the class. Come join our class later when you do have the background knowledge.
Absolute academic integrity is expected of all UBC students in every academic undertaking. UBC gives wide discretion to the course instructor to define what constitutes integrity, i.e., honest and responsible scholarship, in their course.
In this course, you are expected to work collaboratively with others. However, please hand in work that is entirely your own. General discussions with other students are encouraged, and even discussion about certain topics, but what you hand in must have been written by you in its entirety.
You are specifically not allowed to:
Search the internet for existing project ideas and blindly use them without referencing them.
Do not copy and paste someone else's papers. If you like some paragraphs, paraphrase them based on your own understanding of them.
The project in this course is a group project. You are required to work in good faith with your groupmates, to contribute approximately equally, and to completely honestly make explicit what contribution came from whom.
If it is unclear what is allowed and what is not, please ask!
Penalties for violations of academic integrity standards are quite severe and can result in failing the course or being expelled from your program.
Our classroom is a positive space. If you are harassed or made to feel uncomfortable, please bring it to the attention of the instructor immediately. If it is the instructor who is the cause, please immediately bring it to the attention of your departmental advisor or the computer science department chair.