• Time: Thursday, November 8th, 2018 at 5:30pm
  • Place: WEH 8220

Abstract

In this talk I will give an overview on some machine learning problems that can be posed as variational problems. I will look at methods such as k-means, spectral clustering, graph based regression, and neural networks. Often, the limit I will be interested in are as the number of data points goes to infinity. In the graph setting this requires a discrete-to-continuum topology which I will recall. The talk is aimed at a wide audience; I will neither assume knowledge of variational methods or machine learning.

Pizzas will be served.