In optimal transport problems, the aim is to optimally transform a given measure into another one. This has many possible interpretations:
as actual mass transport; as transformations of images; or as transformations of probability distributions. The Data Science and Machine Learning group is starting a reading group on this subject. This week’s seminar will be an overview, followed by discussion and scheduling of future talks. Please consider suggesting papers for the group to read.
Optimal Transport: a Brief Survey with Applications
Richard Blute (University of Ottawa)
In this expository talk, we introduce these algebras as an algebraic way to capture the notion of antidifferentiation. We discuss constructions and examples, and discuss how this may ultimately lead to a theory of integral linear logic.