CSIS Seminar

Algorithmic Fairness: Collisions and Collaborations Between Technology and Society

Abstract

The field of "algorithmic fairness" studies the collisions, and potential collaborations, between a tech-drenched society and issues of justice and equitability. My research over the past 8 years has been about establishing a field whose objective is to understand the societal impact of automation, and especially automation that arises from the data-rich area of machine learning (ML). In this talk I will take you through two vignettes of my research that illustrate the deep and complex interplay between formalisms, algorithms, and broader societal concerns. Firstly, I will discuss the transparency question of knowing why an algorithm makes a particular prediction -- the problem of explanations. I'll outline what we currently know about feature attribution - identifying how much a feature influences a prediction and present my research on the limits of these approaches, applying both mathematical and a broader societal lens. I will also link these insights to a legal consideration of explanations and our ability to foresee harms from ML. Secondly, I'll talk about what it might mean to have a fair clustering. I'll describe definitions and algorithms, and introduce the idea of fair ACCESS --- whether it be about voting rights, access to testing and vaccines, or even access to information in a network - and how we should think about what equity means in this context. I'll conclude with a vision for where research in this area is going next, centered around how we might use technology to support justice and more broadly aid in opportunities for human-flourishing. ................................................................................................................................................................. For in-person attendance: we are following University guidelines so if you are attending the on-campus Distinguished Lecture Series talk, you need to RSVP through the below link: https://docs.google.com/forms/d/e/1FAIpQLSc8hLGh9z--8_CO0VsVPWtMTdmOKLRu472BmEugJ05k2OPJmg/viewform?usp=sf_link Participants must complete Mason COVID Health✓™ and receive a “green light” status on the day of the event. Masks are also required for attendance. ZOOM Link: https://gmu.zoom.us/j/92545949435?pwd=cUVwSjJ0UCsxR1VCWk5KMm9Wcjg3QT09 Meeting ID: 925 4594 9435 Passcode: 007407

Speaker Bio

Suresh Venkatasubramanian is a faculty member in computer science and data science at Brown University, currently on loan to the White House in the Office of Science and Technology Policy. His background is in theoretical computer science, and he’s taken a long and winding path through many areas of data science. For almost the past decade, he’s been interested in algorithmic fairness, and more broadly the impact of automated decision-making systems in society. All views expressed in this talk are solely his own and do not represent the perspectives of any of the institutions with which he is affiliated. ______________________________________________________