One of the course objectives is to learn how to pursue a research problem of your choice and communicate your findings with others. You will have an opportunity to develop those skills in this class by doing a term project. You will form a team of max. 4 students for the term-project. (Optional) A student can work on a project individually if they want, but the expectations will be the same for the team with four members.
You have two options in choosing topics:
- Choose your own topic: You are encouraged to choose any topic relevant to security and privacy of machine learning, but your research problem yet to be explored in the literature. You're welcome to select a topic connected to your current research: for example, if your primary research area is reinforcement learning, you can do a term project on some aspects of security and privacy in reinforcement learning.
- Replicate the results of prior work: You may have a hard time finding a topic of your interest. In this case, you can choose this option and work on reproducing the results in prior work—teams in this track spare time and effort in searching for a topic. To be fair with the first option, I will assign three papers on a topic and ask to reproduce their key results. Please email me for this option.
Please use this Google Sheet [link
] to form a team (by the deadline on syllabus) and sign-up for projects.
Your team should present the research progress three times; I have the following expectations:
Checkpoint Presentation 1 (10 min / team)
- Identify a research problem (or your project scope).
- Do a literature review and explain the prior work relevant to your problem.
- Clearly articulate how your problem is different from the prior work.
[If you choose the second option, you should explain the assigned papers in detail]
- Provide your next steps.
Checkpoint Presentation 2 (10 min / team)
- Design a set of experiments to evaluate your idea's feasibility.
- Obtain preliminary results.
- If you choose your own topic, you must implement a minimal (key) functionality to demonstrate the feasibility of your idea.
- If you choose to reproduce the results of prior work, you must show at least the key results of one of the papers provided by me.
- Provide your next steps.
Final Presentation (15 min / team)
- Deliver your final results (based on the goals you set in Checkpoint 1).
- In-class presentation and a report [Note: you can submit your report by the 14th].
Grading Policy (Evaluations)
Your term project will be evaluated with the following scheme (35 in total):
[Peer reviews; 5 points for each presentation]
- 10 points: Checkpoint Presentation 1
- 10 points: Checkpoint Presentation 2
- 15 points: Final Presentation and Write-up (report template)
In each checkpoint, you will be asked to evaluate other team's progress (2-3 teams other than yours). I will provide a rubric for the peer review; you will also be asked to provide constructive feedback to those teams. Note that I will collect the reviews; they will be sent to each team anonymously.