Скачать или смотреть On Parameterizing Optimal Transport with Elastic Costs with Marco Cuturi

In this video, Marco Cuturi from Apple, France, provides an in-depth overview of optimal transport (OT) computations, with a focus on the challenge of computing OT maps using two samples from high-dimensional probability measures. After reviewing several popular methods for this task, including those that utilize neural architectures, Cuturi introduces his recent work on parameterizing OT problems with elastic costs. These costs combine classic squared Euclidean distance with regularizers like the L1 norm. He’ll explain the properties of OT maps that follow such costs, and present new methods for both computing ground truth OT maps with elastic costs and adaptively learning the parameters of the regularizer.

This talk took place on Thursday, October 24 at the 2024 SIAM Conference on Mathematics of Data Science.

Learn more about SIAM https://www.siam.org/
Join SIAM https://www.siam.org/membership/indiv...
Attend a SIAM Conference https://www.siam.org/conferences-events

#apple #optimization #transportation #probability #neuralnetworks #datascience #dataanalytics #euclidean #machinelearning #mathematics #mathematicseducation