Mathematics Colloquium
Rocio Diaz Martin
Tufts University
Title: Metrics Based on Optimal Transport
Date: Monday, January 13, 2025
Place and Time: Love 101, 3:05-3:55 pm
Abstract. The optimal transport (OT) problem seeks the most efficient way to transport a distribution of mass from one configuration to another, minimizing the associated transportation cost. This framework has found diverse applications in machine learning due to its ability to define meaningful distances, known as Wasserstein distances, between probability distributions. However, Wasserstein distances can be computationally expensive, particularly in high-dimensional settings. In this talk, we will introduce new OT-based metrics designed to address these computational challenges. These metrics preserve the desirable properties of the Wasserstein distance while offering significant improvements in efficiency, making them more suitable for large-scale applications.