Fairness, Representation, and Machine Learning
Overview
This assignment builds on introductory knowledge of machine learning techniques, namely the naïve Bayes algorithm and logistic regression, to introduce concepts and definitions of algorithmic fairness. Students analyze sources of bias in algorithmic systems, then learn formal definitions of algorithmic fairness such as independence, separation, and fairness through awareness or unawareness. They are also introduced to notions of fairness that complicate the formal paradigms, including intersectionality and subgroup analysis, representation, and justice beyond distribution.
Contributors
- Ethics materials by Kathleen Creel.
- Problem Set by Chris Piech, Jerry Cain, Kathleen Creel, Derek McCreight, and Tim Gianitsos
Assignment goals
- Learn and apply formal definitions of fairness, many of which rely on probabilistic concepts.
Ethics goals
- Understanding formal definitions of algorithmic fairness.
- Exploring the limits of formal fairness and contrasting it with other normative concepts such as representation and justice.
Download Links
- Problem Set or Exam Questions (pdf or tex )
- Ethics Slides on Fairness and Representation (pptx)
- Explainer Handout
Additional Readings for Context (Instructors or Students):
- ACM Code of Ethics and Professional Conduct
- Srinivasan & Uchino, Biases in Generative Art: A Causal Look From the Lens of Art History
- Friedman & Nissenbaum, Bias In Computer Systems
- hooks, In Our Glory: Photography and Black Life
- Dwork, Fairness Through Awareness
- Sapiezynski, Ghosh, & Kaplan, Algorithms that “Don’t See Color”: Comparing Biases in Lookalike and Special Ad Audiences
- Barocas & Selbst, Big Data’s Disparate Impact
- Barocas, Hardt, Narayanan, Fairness and Machine Learning: Limitations and Opportunities