Course Home Page for Ethical Algorithm Design (CIS 423/523) University of Pennsylvania, Spring 2022 (original) (raw)
Ethical Algorithm Design CIS 423/523
Spring 2022
Tuesdays and Thursdays 1:45-3:15PM ET
118 Fagin Hall
Instructor:
Prof. Michael Kearns
mkearns@cis.upenn.edu
Office hours (virtual): TBD
Teaching Assistants:
Ira Globus-Harris (they/them)
igh@seas.upenn.edu
Office hours: Virtual Wed 11-1hereor by appointment.
Elizabeth Margolin
ecmargo@seas.upenn.edu
Office hours: Virtual Thu 11-1hereor by appointment.
Course Description
This course is about the social and human problems that can arise from algorithms, AI and machine learning, and how we might design these technologies to be "better behaved" in the first place. It is first and foremost a science or engineering course, since we will be developing algorithm design principles. You can get a rough sense of course themes and topics by visiting the websites for the pilot versions of this course offered in2021, 2020and2019
Here are the lecture videosfrom the pilot version from last year. Please note that they will not correspond exactly to this year's lectures, and should not be viewed as a substitute for mandatory lecture attendance.
Prerequisites: Familiarity with some machine learning, basic statistics and probability theory will be helpful. While this is not a theory class, you need to be comfortable with mathematical notation and formalism. There will be some simple coding and data analysis assignments, so some basic programming ability is needed.
Course content will include readings from the scientifc literature, the mainstream media and other articles and books.
Grades will be based on some TBD mixture of quizzes, coding assignments, possible a midterm and/or final, and perhaps a final project or book club.
Since at least the first couple of lectures will be virtual, here is theZoom linkfor the live sessions.
CIS 423/523 fulfills theSEAS Engineering Ethics Requirementfor these programs: ASCS, BE, CMPE, CSCI, DMD and NETS.
| Lecture Dates | Topic | Slides, Readings, Assignments, Announcements |
|---|---|---|
| Thu Jan 13 | Course Introduction and Overview | Lecture video:[Jan 13] Lecture Notes The first assignment of the semester is to completethis survey. A general-audience introduction to some of the themes of the course is given in the (recommended but not required) bookThe Ethical Algorithm: The Science of Socially Aware Algorithm Design,by M. Kearns and A. Roth.Also recommended but not required:Weapons of Math Destruction: How Big Data Increases Inequality and Threatens Democracy,by C. O'Neil. |
| Tue Jan 18Tue Jan 25 | Foundations of Machine Learning | Lecture videos:[Jan 18] Lecture Notes There is good and comprehensive set of videos and readings related to many of the topics we covered in these lectures in thisGoogle machine learning course. |
| Thu Jan 27Tue Feb 1 | Discrimination in Machine Learning: COMPAS and ProPublica | Lecture Notes The following readings are required; you should read the two ProPublica pieces before the Jan 25 lecture so we can discuss them then. ProPublica article on COMPAS ProPublica analysis COMPAS Risk Assesment Survey(just skim) Northpointe response to ProPublica ProPublica github repository, including dataset(we'll look at the dataset a bit in lecture)(technical, just skim) |
| Thu Feb 3Tue Feb 8Thu Feb 10Tue Feb 15Thu Feb 17Tue Feb 22Thu Feb 24 | Science of Fair ML: Models and Algorithms | Lecture Notes Readings: Inherent Trade-Offs in the Fair Determination of Risk Scores,J. Kleinberg, S. Mullainathan, M. Raghavan. (Read first 8 pages) The Frontiers of Fairness in Machine Learning.Alexandra Chouldechova, Aaron Roth. (Read entire article)Please play around with the followingGoogle demo site on fairness and ML. Equality of Opportunity in Supervised Learning.M. Hardt, E. Price, N. Srebro. (Optional; this is the post-processing/bolt-on method.) A Reductions Approach to Fair Classification.A. Agarwal, A. Beygelzimer, M. DudÃk, J. Langford, H. Wallach. (Optional; this is the in-processing/constrained optimization/game theory method.)And here is a series of papers extending the game-theoretic approach to different/stronger notions of fairness (all optional): An Empirical Study of Rich Subgroup Fairness for Machine Learning.MK, S. Neel, A. Roth, S. Wu. Average Individual Fairness: Algorithms, Generalization and Experiments.MK, A. Roth, S. Sharifi-Malvajerdi. An Algorithmic Framework for Fairness Elicitation.C. Jung, MK, S. Neel, A. Roth, L. Stapleton, S. Wu. Minimax Group Fairness: Algorithms and ExperimentsEmily Diana, Wesley Gill, MK, Krishnaram Kenthapadi, Aaron Roth. |
| Tue Mar 1Thu Mar 3 | "Bias Bounty" Project Introduction | Here is the algorithmic ``bias bounty'' framework; it is recommended that you take at least a high-level reading for the coding project: Beyond the Frontier: Fairness Without Accuracy Loss.I. Globus-Harris, MK, A. Roth. |
| Tue Mar 15Thu Mar 17Tue Mar 20Thu Mar 22 | Algorithmic Privacy and Related Topics | Lecture Notes |