Scaling Optimal Transport for High Dimensional Learning
Wednesday April 12, 2023 - 14.30 (GMT+1)
Gabriel Peyré
École normale supérieure
Embedded Files
Scaling Optimal Transport for High dimensional LearningAbstract: Optimal transport (OT) has recently gained a lot of interest inmachine learning. It is a natural tool to compare in a geometricallyfaithful way probability distributions. It finds applications in bothsupervised learning (using geometric loss functions) and unsupervisedlearning (to perform generative model fitting). OT is however plagued bythe curse of dimensionality, since it might require a number of sampleswhich grows exponentially with the dimension. In this talk, I will explainhow to leverage entropic regularization methods to define computationallyefficient loss functions, approximating OT with a better sample complexity.More information and references can be found on the website of our book"Computational Optimal Transport" https://optimaltransport.github.io/
Page updated
Google Sites
Report abuse