Robustness and sensitivity to outliers

Professor: Robert Webber

Description: In data science, an important research area is sensitivity to outliers — should a model depend on the most extreme data points, or only the typical ones? My project explores the sensitivity to outliers problem through implementing and analyzing optimization problems in clustering, nonnegative matrix factorization, and column subset selection.

Preferred Qualifications: Strong coursework in probability and linear algebra is highly valued.