Class 25: Privacy and Incentives

Schedule Update

Thursday’s class will be project review meetings. We will meet with each of the teams to discuss feedback on your proposals and progress on the projects. Weather permitting, class will meet at the picnic tables in the Thornton hall courtyard. Please ensure that your team is ready to discuss issues raised in the feedback we’ve sent on your project, explain your progress so far, and any things you think we can help with.

Slides

Download (full resolution) PDF


Class 24: Privacy

Schedule

Your Final Project Proposal is due tonight at 8:59pm.

Slides



Download (full resolution) PDF

Links

Stanley Warner. Randomized Response: A Survey Technique for Eliminating Evasive Answer Bias, Journal of the American Statistical Association, March 1965.

Cynthia Dwork and Aaron Roth. The Algorithmic Foundations of Differential Privacy. NOW Foundations and Trends in Theoretical Computer Science, 2014.

Burton H. Bloom. Space/Time Trade-offs in Hash Coding with Allowable Errors, Communications of the ACM, July 1970.

Ăšlfar Erlingsson, Vasyl Pihur, Aleksandra Korolova. RAPPOR: Randomized Aggregatable Privacy-Preserving Ordinal Response. ACM CCS 2014. Code: https://github.com/google/rappor.

Privacy-Preserving Machine Learning.
Bargav Jayaraman and David Evans. When Relaxations Go Bad: “Differentially-Private” Machine Learning. February 2019.


Class 21: Economics of Information

Slides

Download (full resolution) PDF


Final Project: Your Choice

For the Final Project, it is up to you to decide what to do, consistent with the course goals.

Teams

You may work in a team of from 1 to 38 students, but we expect most teams to be 2-4 students. If you want to work alone, you will need to make a convincing case why you will get more from the project by working alone than by working with others.

The impressiveness of your project should scale with at least the square root of the number of class participants on your project team. If you want to form a team with more than four students, you will need to make a case that you have a project idea that benefits from such a large team, and a management plan for making a large team successful. There are no other constraints on the formation of your team, although we expect most teams will be a mix of CS and Economics majors.

Regardless of the size of your team, you should have a project plan that involves contributions from all team members. Everyone on the team should have a clear responsibility for parts of the project, and contribute substantially to the overall success of the project. Except in extraordinary circumstances, all team members will receive the same grade.

Topic

For your project topic, you can work on any topic you would like that is relevant to the course. This means it should involve aspects of both Economics and Computer Science, with broad definitions of what each discipline encompasses.

A project topic should satisfy at least three of these goals:

  • fun (for you to do, and for others to see)
  • relevant (to the class)
  • technically interesting
  • useful (at least to your team, but hopefully to many)

We will provide some ideas for potential projects to help you think of a project idea, but you should not feel constrained by these suggestions, and we hope many student teams will come up with creative ideas that are not based on our suggestions.

Deliverables

There are four deliverables for the Final Project, described next.

Project Proposal

Due Thursday, 11 April (8:59pm). Your project proposal should contain:

  1. Title of your project: short description that clearly captures your project idea.

  2. A short paragraph that describes the goal of your project.

  3. A project jusfication that explains for at least three of the goals above (fun, relevant, technically interesting, useful) how your project satisfies them.

  4. A project plan that explains the main tasks needed to successfully complete your project and what you will actually do.

  5. Resources you have found or your plan for finding them. For most projects, this should include a list of papers relevant to your project. For many projects, it will also include datasets and code that you plan to use.

  6. A list of your team members and their roles and responsibilities. If your team has more than two people, this should also explain how you plan to coordinate and manage your team.

You should submit your project proposal by sending a slack message to a group that includes all of your team members and all of the course staff: @Dave @Denis Nekipelov @Joe Roessler @Jonas. You can submit the proposal as a PDF file attachment to the message.

Project Team Meetings and Progress Reports

These will be scheduled individually, the week of April 15. All team members are expected to participate in the team meeting. At the team meeting, you will be expected to adress any questions that were raised about your proposal, and to explain what your team has done so far. This is also an opportunity to ask any questions you have and get advice from the course staff.

Project Presentations

In class, on Tuesday, 30 April (the last day of class), each team will have an opportunity to present your project. (Details will be announced later.)

Project Reports

Final project reports are due, Monday, 6 May (4:59pm). The format of the report will depend on your topic and how to best present it, but we hope many teams will end up with project reports that are interactive web sites (which could be built from Jupyter notebooks) and include open source code and data.


Final Auction Results

Here are the final auction results:

Team Clicks Bids Submitted Budget Unspent
4507 53 3274 $0.42
4501 48 3380 $0.68
4509 45 3094 $222.04
4504 41 3362 $0.06
4506 37 1986 $0.94
4502 37 3487 $0.63
4505 26 631 $2,965.38
4503 22 1658 $0.40
4510 21 3450 $0.12
4508 13 694 $2.82

Congratulations to Team 7: Cyrus Morshedi, Ian Hardman, Ryan Dailey, and Hunter Rockley!