Faculty Candidate Interview

[mikeizbicki]

The CMC Mathematical Sciences Department invites you to Faculty Search Candidate Talks:

Featuring: R. Teal Witter, New York University, https://www.rtealwitter.com/

Research Talk: Algorithmic Foundations of Trustworthy Machine Learning 12/05 (Thursday) 4pm-5pm, Davidson, Adams Hall Teaching Demo 12/06 (Friday) 11:00am-11:50am, Kravis 165

Title: Algorithmic Foundations of Trustworthy Machine Learning

Abstract: While modern machine learning models can achieve impressive results, we often don't understand how they make their decisions. This lack of transparency is particularly concerning when these systems are deployed in high-stakes domains like healthcare, law, and finance, where we need to trust and verify automated decisions. My research focuses on developing theoretically sound approaches to make machine learning more interpretable and trustworthy. In this talk, I'll discuss my work on creating rigorous algorithms that explain how machine learning models arrive at their predictions. I'll focus on recent progress in efficiently computing Shapley values—a concept borrowed from game theory that helps us understand how different inputs contribute to a model's decisions. We'll discuss a recent algorithm that leverages ideas from randomized linear algebra to provide mathematical guarantees about its accuracy, and state-of-the-art performance. This work demonstrates how theoretical computer science can help solve practical challenges in modern artificial intelligence, making these powerful technologies more reliable and trustworthy for real-world applications.

[mikeizbicki]

You can also have the opportunity to eat lunch with our faculty candidates. Teal Witter will have lunch with students 12-1pm on Friday 6 Dec.

The other candidates are scheduled for Monday 9 Dec from 12-1pm and Wednesday 11 Dec 12-1pm.

All the lunches will be at Collins. (We will pay.) The first lunch will have Prof. Sarah Cannon walking the candidate down, so you can look for her.