Randomized Algorithms for Data Science
In this class, we will discover how data science techniques are deployed at scale. The questions we investigate will include: How do services such as Shazam recognize song clips in seconds? In settings with hundreds of features, how do we find patterns? Given a social network, how can we detect groups? And how can we use vibrations to "see" into the earth? We'll answer these questions and more by exploring how randomization lets us get away with far fewer resources than we'd otherwise need. Topics include random variables, concentration inequalities, dimensionality reduction, singular value decomposition, spectral graph theory, and approximate linear regression. (MATH 0200, CSCI 0200 and CSCI 0302)

Teal Witter is a PhD candidate at NYU Tandon. He graduated from Middlebury in 2020 and can't wait to return to snowy Vermont for the winter term!/