2026-27 J.C. Eaves Lecture
Topic: Decision boundaries: How they help us understand algorithms from classical ML to modern AI.
Decision boundaries are geometric patterns that indicate the ways classifiers label data points. Most people in data science and machine learning have seen them, but thinking deeply about them helps us understand more clearly the behaviors of the different classical classifier algorithms, including logistic regression, decision tree, random forestd and k-NN. They also help us think about what's going on inside modern AI algorithms such as chatbots. In this talk, I'll show how decision boundaries are an overlooked topic that help us better understand the old and the new in machine learning.
Bio: Noah Giansiracusa is a math professor at Bentley University (a business school near Boston) and a faculty associate at Harvard. He’s appeared on CNN, NBC, and the BBC and has written for a range of outlets including Time, Scientific American and the Washington Post. He’s written two books: "How Algorithms Create and Prevent Fake News" (2021) and "Robin Hood Math: Take Control of the Algorithms that Run Your Life" (2025). His work has been praised by multiple Nobel laureates, an Ironman champion and the most powerful woman in hedge funds, among others.



About the speaker: 
About the speaker: Mihai Stoiciu received his undergraduate degree in Mathematics at the University of Bucharest, Romania and his Ph.D. (also in Mathematics) at Caltech in Pasadena, CA (dissertation title: "Zeros of Random Orthogonal Polynomials on the Unit Circle"). After graduating from Caltech, Stoiciu started as an Assistant Professor at Williams College, where he is now an Associate Professor. At Williams, Stoiciu teaches courses at all levels and has advised seven Honors Theses and three summer REU Programs. He spent leaves and sabbaticals at the Newton Institute for Mathematical Sciences in Cambridge, UK, at University of California Irvine, and at University of Wisconsin Madison. Stoiciu's research is in Mathematical Physics and Functional Analysis and his current interests are in Schrodinger operators and random matrices.