Assistant Prof at Wellesley College CS | Research in algorithms, fairness, and democracy | Board game designer | Former filmmaker | vegetarian | he/him
On my way to #Unpub Prime for the first time since the pandemic. I'll be there every day and have a table Sunday. Looking forward to seeing old friends and meeting new ones! I'm bringing a big worker placement game about ants, a few 18-card microgames, and more. @TheUnpub
I’ll be at #AAAI2024@RealAAAI this week to present our paper, “Implications of Distance over Redistricting Maps: Central and Outlier Maps”. Looking forward to catching up with friends and meeting folks interested in algorithmic fairness for ML, OR, and democratic systems.
This work with Seyed A Esmaeili, Darshan Chakrabarti, and Hayley Grape, presents a new way to look at the problem of congressional redistricting and gerrymandering. We introduce a distance measure over maps that defines a central "typical" map within an ensemble.
Kristian Lum(@KLdivergence), a renowned statistician and machine learning researcher who focuses on fairness, accountability and transparency, has joined the Data Science Institute as a research associate professor.
Learn more about Kristian here: https://t.co/zoMGwKiQqm
I'm currently on the job market for industry & academic research positions! My research leverages ethnography to understand how frontline healthcare workers perceive AI & examines how XAI can be designed for emerging technology users.
Happy to discuss all relevant opportunities!
@KLdivergence Yeah. The minions you expect, but the insensitivity of those memes was disappointing. I’m glad you all got to be part of that great team while it lasted and it’s nice to see the heartfelt messages about it today. Thinking of all of you and where the future lies for this work.
@gengelstein I only skimmed the article, but this looks a little too close to, "We trained a chatbot on fringe political parties. What could go wrong?"
It’s a little weird that the COVID close contact tracker app for academic conferences is Twitter. So far, I’m feeling okay after #EAAMO2022. Slightly jealous of some of the great work I saw there, but no sore throat.
#EAAMO2022 was great! Thanks to @ACMEAAMO organizers and all the volunteers. I highly recommend this conference which brings together a great group of people doing really interesting and fresh work. My first in-person conference since the pandemic and it was very energizing.
#EAAMO2022 is officially at an end! Our infinite thanks to all our speakers, attendees and volunteers for making our first in-person conference a big success. 🙌 It has been amazing to hear about the diverse and inspiring work being done by our community. (1/3)
@gengelstein Cool. I’ve been thinking about ways to add uncertainty that children are able to understand enough to make predictions. Maybe the visual features of asymmetric dice could support this. The current gold standard for me is @HABAGamesUSA’s Dragon’s Breath. Kids just get it.
@kiragoldner@Tim_Roughgarden I still see CS papers on drawing equitable public school districts as if people won't move or switch to private school. I touched on this at the end of our fair clustering tutorial (https://t.co/TUcWdKuW15) & found a great Thurgood Marshall quote from Milliken v. Bradley (1974).
Of course, there are plenty of situations in which RCV improves over the plurality system most states use, but it doesn’t seem to be having a major effect in the Alaska special election.
People shouldn’t view the outcome in Alaska as indicating whether RCV is good or bad. n/n
I wish coverage of the Alaska special election would explain clearly that the switch to ranked choice voting (RCV) isn’t really affecting the outcome. 🧵1/n
New Alaska count: Mary Peltola (D) expands lead over Sarah Palin (R) to 16,347 votes. Depending on how many ballots “exhaust,” Palin probably needs 68-70% of Begich second place votes to prevail. It’s going to be close. #AKAL
Paradoxically, RCV could have had an effect though if the state skewed more Republican.
A 75/25 party split could cause RCV to favor a less far-right republican like Begich. If Begich got more votes than Peltola, he could have gotten most of her 2nd choice votes and won. 3/n