SMSS Colloquium Talk - Leland McInnes

Speaker: Leland McInnes, Tutte Institute, Ottawa

Leland McInnes completed his PhD in 2007, under the supervision of David Riley in the Department of Mathematics at Western. Dr. McInnes is a research mathematician at the Tutte Institute.

Title: Topological Data Analysis for Dimension Reduction and Clustering

Abstract:

Topological Data Analysis is a relatively new field of research that seeks to bring powerful tools from algebraic topology to bear on practical problems in data science (where statistics has been the more traditional toolset). In this talk we will looks at a new approach to dimension reduction called UMAP, that makes use of tools from algebraic topology to provide better practical results. We'll also look at how the mathematical ideas behind this approach can provide a more general lens through which to view unsupervised learning, and propose a new clustering algorithm built using the same tools and techniques. This approach potentially offers a new mathematical approach and novel new semantics for clustering.