BIMSA >
Seminar on Control Theory and Nonlinear Filtering
Geometric Methods of Machine Learning (I): Linear Methods
Geometric Methods of Machine Learning (I): Linear Methods
Organizer
Speaker
Time
Friday, September 29, 2023 8:30 PM - 9:00 PM
Venue
Online
Abstract
This and the next two presentations is a summary of the book Mathematical Principles of Topological and Geometric Data Analysis. In recent years, we have seen the fascinating development that concepts and methods developed in the context of research in pure mathematics, and in particular in geometry, have become powerful tools in machine learning for the analysis of data. In this book, we shall first develop the background from topology, geometry, measure theory, graph theory in Chaps. 2-5 and then systematically present and analyze geometric methods of machine learning in Chaps. 6, 7 and 8. This presentation is about Chap 6, which discusses linear methods, including kernels, variations, and averaging.