Ira Globus-Harris
Pronouns: they/them
PhD Student, Computer and Information Sciences
University of Pennsylvania
Email: igh at seas dot upenn dot edu

About Me

My name is Ira Globus-Harris, and I am a fifth year PhD student in the computer and information sciences department at the University of Pennsylvania, co-advised by Aaron Roth and Michael Kearns. My work, which is grounded in the algorithmic foundations of responsible computing, broadly looks at mechanisms to resolve potential harms incurred by AI-driven decision-making. In particular, my work focuses on using algorithmic techniques which are scalable for real-world use, holistically consider an algorithm in its broader context, and which flexibly incorporate human input. I am particularly interested in methods which provide flexibility to adjust models after deployment in scalable and efficient ways.

Prior to Penn, I worked at Boston University as a software engineer with SAIL and with the Harvard Privacy Tools Project. I got my bachelors degree from Reed College in Portland, OR.

Publications

Globus-Harris, Ira; Gupta, Varun; Kearns, Michael; Roth, Aaron. "Ensembing for Constrained Optimization". 2024.

Globus-Harris, Ira; Harrison, Declan; Kearns, Michael; Perona, Pietro; Roth, Aaron. " Diversified Ensembling: An Experiment in Crowdsourced Machine Learning. " 2023.

Globus-Harris, Ira; Harrison, Declan; Kearns, Michael; Roth, Aaron; Sorrell, Jessica. "Multicalibration as Boosting for Regression." ICML 2023 (Oral presentation).

Globus-Harris, Ira; Gupta, Varun; Jung, Christopher; Kearns, Michael; Morgenstern, Jamie; Roth, Aaron. " Multicalibrated Regression for Downstream Fairness". AIES 2023.

Globus-Harris, Ira; Kearns, Michael; Roth, Aaron. "An Algorithmic Framework for Bias Bounties." ACM Conference on Fairness, Accountability, and Transparency (FAccT) 2022.

Diana, Emily; Gill, Wesley; Globus-Harris, Ira; Kearns, Michael; Roth, Aaron; Sharifi-Malvajerdi, Saeed. "Lexicographically Fair Learning: Algorithms and Generalization." Foundations on Responsible Computing (FORC). 2021.

Drechsler, Joerg; Globus-Harris, Ira; McMillan, Audra; Sarathy, Jayshree; Smith, Adam. “Non-parametric Differentially Private Confidence Intervals for the Median.” 2021.

Swanberg, Marika; Globus-Harris, Ira; Griffith, Iris; Ritz, Anna; Groce, Adam; Bray, Andrew. “Improved Differentially Private Analysis of Variance.” Privacy Enhancing Technologies Symposium (PETS). 2019.

Qin, Lucy; Lapets, Andre; Jansen, Frederick; Flockhart, Peter; Dak Albab, Kinan, Globus-Harris, Ira. "From Usability to Secure Computing and Back Again.” Fifteenth Symposium on Usable Privacy and Security (SOUPS). 2019

Dak Albab, Kinan; Issa, Rawane; Lapets, Andrei; Flockhart, Peter; Qin, Lucy; Globus-Harris, Ira. “Tutorial: Deploying Secure Multi-Party Computation on the Web Using JIFF.” IEEE Secure Development (SecDev), 2019.

Talks

"Diversified Ensembling: An Experiment in Crowdsourced Machine Learning," Oral presentation at FAccT, June 2024. "Collaborative Crowdsourced Machine Learning," Stanford Fairness Seminar, May 2024. "Multicalibration as Boosting for Regression," Institute for Operations Research and Management Sciences (INFORMS) Annual Meeting, October 2023.
"Multicalibration as Boosting for Regression," Oral presentation at ICML, August 2023.
"Multicalibration as Boosting for Regression," Foundations on Responsible Computing (FORC), June 2023.
“An Algorithmic Framework for Bias Bounties,” Institute for Operations Research and Management Sciences (INFORMS) Annual Meeting, October 2022.
“An Algorithmic Framework for Bias Bounties,”Oral presentation at FAccT, June 2022.
“An Algorithmic Framework for Bias Bounties,” Reed College Computer Science Colloquium, April 4 2022.
"Lexicographically Fair Learning: Algorithms and Generalization," Theory of Computation for Fairness Seminar, Simons Institute Collaboration, October 27 2021.

Awards and Recognitions

AWS AI ASSET Fellow, 2024.
EECS Rising Star by Georgia Tech, 2023.

Teaching

Fall 2022-Spring 2024: Undergraduate thesis advisor for Reed College
Spring 2022: Head TA and course design, U Penn CIS 423/523 Ethical Algorithm Design
Spring 2022: TA, U Penn NETS 412 Algorithmic Game Theory
Fall 2021: Head TA and course design, U Penn CIS 320 Introduction to Algorithms
2015-2018: TA and tutor, Reed Math 121 Introduction to Programming
2017-2018: Instructor and course design, Computer Science Outreach Program
2017: Tutor, Reed Physics 202 Modern Physics
2016: Tutor, Reed Physics 201 Oscillations and Waves
2015-2016: Tutor, Reed Humanities 110

Past Research Affiliations and Projects

Amazon AWS intern (Summer 2021-Spring 2022, Summer 2023, Summer 2024)
Boston University SAIL
Harvard Privacy Tools Project
SmartNoise
OpenDP
Multiparty.org

Advocacy

I was the president of the Penn school of engineering and applied sciences' group for LGBTQ+ graduate students from 2020-22, and am a diversity council representative for the school of engineering and applied sciences. While at Penn, I have led and advocated for several changes for trans inclusion within the school and the university more broadly. While in undergrad, I co-led the group for gender minorities in STEM fields. I also designed curriculum and taught a computer science outreach program to a local elementary school while at Reed, and painted some fun murals which have now inspired some other fun murals.