Instructions

• This is not role-playing; you're just a student teaching the other students. But you are responsible not just for learning about this yourself but teaching this material to your fellow students! Take your job seriously for their benefit as well.
• This will sometimes require reading multiple papers to get enough context, and even if it says "skim" or "abstract" in the assignment, if you are assigned this topic you obviously have to read them in full and carefully. (The links are in the pages for the relevant weeks.)
• You are being paired up partly so you have someone to work with and bounce ideas off but also so you can debate what you think about the reading, separately from simply completing the assignment. Take some time to step back and reflect on your own take on whatever the topic is.
• You should expect to draw on the readings from previous weeks and feel free to briefly recap relevant material in the presentation to remind the audience if it is helpful (ML concepts, fairness, privacy, etc). The first part of the semester was about building a foundation for reading these and you should reread some of the more abstract/theoretical articles from previous weeks as needed, in the context of the specific topic you are assigned and try to connect it back to the core concepts. That will help in making the ideas concrete both for you and the other students. For example, if your topic is an analysis of fairness issues in a public sector system, the paper "Fairness and Accountability Design Needs for Algorithmic Support in High-Stakes Public Sector Decision-Making" that you read in the digital rights + fairness week is something you should probably revisit, as are the basic technical readings of definitions of algorithmic fairness.
• Feel free to Google around to get more context on the topic you are assigned; this is not any kind of "closed book" exam. If you happen to find commentary on your exact topic (e.g., NYPD surveillance systems), do not just copy someone else's analysis — expect to critique their analysis as well or just use it to give you a little context for how other people (who may in some cases actually know less than you already do) may be thinking about it.
• Submit PDF slides to Liam via email by noon on the day you are presenting so they can be combined before each class to avoid wasting time. I will not grade the slides separately, just the overall presentation.
• You will have 20 minutes overall, but your presentation should be around 10 minutes to allow for class discussion + Q&A after.
• This piece with advice on giving talks may be useful (though note it is targeted more at presenting your own research rather than explaining someone else’s).

March 19

Team #1 (Finance) - Alice Frank & Hirofumi Honzawa

S. Trilling, Fair Algorithmic Housing Loans, Aspen Tech Policy Hub, 2020

Consider the other readings in this section complementary.

Team #3 (Healthcare) - Magan Chin, Samantha Alonso, Yunis Gurbanov

Z. Obermeyer, B. Powers, C. Vogeli, and S. Mullainathan, Dissecting racial bias in an algorithm used to manage the health of populations, Science, 2019

March 26

Team #4 (Crime) - Jeremie Ponak & Hamna Khan

• J. Angwin, J. Larson, S. Mattu, and L. Kirchner, Machine Bias, ProPublica, 2016
• J. Larson, S. Mattu, L. Kirchner, and J. Angwin, How We Analyzed the COMPAS Recidivism Algorithm, ProPublica, 2016

Team #5 (Crime) - Kendal Gee, Lauren Goldberg, Reagan Henry