Tropical Geometry Tools for Machine Learning and Phylogenomics

Linear algebra provides key tools in statistical and machine learning models. Tropical geometry can be described as the piecewise-linear analogue of classical geometry and in this talk we discuss recent exciting developments in “tropical geometric machine learning tools” in the rapidly growing area of phylogenomics, which can provide insight into relationships and evolutionary patterns of a diversity of organisms, from humans, plants and animals, to microbes and viruses. This is joint work with a subset of the following wonderful researchers: David Barnhill, George Aliatimis, John Sabol, Chris Teska, Keiji Miura, Kurt Pasque, Jefferson Huang.