Optimal transportation theory has been proved versatile and useful for a bunch of applications in economics, social sciences (e.g. crowd motion), structured data analysis (e.g. image processing), PDE's in Physics and Engineering (e.g. fluid dynamics, optics), Operations Research and many others. Several discretization methods and numerical tools are available to make optimal transportation an effective paradigm for its many applications also from a computational point of view.
In this course, after general introductory material and primal/dual formulation of the underlying optimization problem, several popular discretization methods will be presented as well as many algorithms, with an emphasis also on their computational complexity. The overall goal will be to give a precise intuition of the underlying mechnisms, relying on rigorous proofs only in simple cases. Some applications in a "hands on" session will be presented and implemented in python and/or jupyter ambient
Tentative schedule (subject to time slot adjustments)
Mon 16 May 9:30-11:30 Aula G
Tue 17 May 8:30-10:30 Aula G
Wed 18 May 13:30-15:30 Aula G
Thu 19 May 16:30-18:30 Aula G/Cyberphysical Lab
Contact person: giandomenico.orlandi[at]univr.it