Past Event: Oden Institute Seminar

Archetypal Analysis

Braxton Osting, Associate Professor, Department of Mathematics, University of Utah

Abstract

**This seminiar will be presented live in POB 6.304 and via Zoom.**

Archetypal Analysis is an unsupervised learning method that uses a low-dimensional convex polytope to summarize multivariate data. For fixed k, the method finds a convex polytope with k vertices, called archetype points, such that the polytope is contained in the convex hull of the data and the mean squared distance between the data and the polytope is minimal. In this talk, I'll give an overview of Archetypal Analysis and discuss our recent results on consistency, a probabilistic method for approximate archetypal analysis, a version of the problem using the Wasserstein metric, and an application in analyzing online hate speech. Parts of this work are joint with Katy Craig, Ruijian Han, Richard Medina, Lam Nguyen, Tanner Sims, Dong Wang, Yiming Xu, Ryeongkyung (RK) Yoon, and Dominique Zosso.

Biography

Braxton Osting is an Associate Professor in the Department of Mathematics at the University of Utah. Prior to moving to Utah, he earned a Ph.D. in Applied Mathematics at Columbia University and was an NSF Postdoctoral Fellow in the Department of Mathematics at the University of California, Los Angeles.  He has broad interests in analytical and computational methods for problems in applied mathematics, especially in partial differential equations, optimization and control, graph theory, and machine learning.