by New faculty member Gavin Brown joins UW-Madison from a postdoc position at the University of Washington. In his research, he focuses on the question, “What happens when the models we train or the analyses we run leak personal information?”
Hometown:
Mount Gilead, Ohio
Educational/professional background:
I did my undergraduate work at Case Western Reserve University and went to graduate school at Boston University. Before moving to Wisconsin, I was a postdoctoral scholar at the University of Washington.
What are your areas of focus?
I work on the problem of learning from data, with a focus on questions of privacy: What happens when the models we train or the analyses we run leak personal information?
What main issue do you address or problem do you seek to solve in your work?
I spend a lot of time designing new techniques to preserve privacy.
Please describe your work for people without a background in computer science:
Much of my work uses differential privacy, a framework for giving mathematical guarantees about privacy in data analysis. Differential privacy was used for the 2020 US Census and is deployed by companies like Apple and Google. I work to improve this technology and invent new techniques, with the goal of facilitating privacy-preserving learning for a wide range of applications.
What’s one thing you hope students who take a class with you will come away with?
I hope my students appreciate the power of even the most modest modern computers.