Deep kernel-based distances between distributions

Abstract

Optimal transport, while widespread and effective, is not the only game in town for comparing high-dimensional distributions. This talk will cover a set of related distances based on kernel methods, in particular the maximum mean discrepancy, and especially their use with learned kernels defined by deep networks. This set of distance metrics allows for effective use in a variety of applications; we will cover foundational properties and develop variants useful for distinguishing distributions, training generative models, and other machine learning applications.

Date
2021-01-30 9:00 AM PST
Speaker
Danica Sutherland (University of British Columbia, CS)
Location
Online (zoom)
Pacific Institute for the Mathematical Sciences

This event is part of the Pacific Interdisciplinary Hub on Optimal Transport (PIHOT) which is a collaborative research group (CRG) of the Pacific Institute for the Mathematical Sciences (PIMS).