Optimal Transport

The talk can be watched here.

What is the most efficient way to transport and modify a pile of sand into another one?
This is the question that prompted brilliant scientists like Monge, Kantorovich, Villani and Figali to investigate the theory of transport between two distributions. Today, optimal transport is an invaluable theory for comparing probability distributions, and it yields countless applications in computer vision or machine learning.

• Aude Genevay is a postdoc in the Geometric Data Processing group at MIT. Her research is focused on the interaction between optimal transport and machine learning. She earned her PhD on Entropy-Regularized Optimal Transport for Machine Learning under the direction of Gabriel Peyré.
• Anton Mallasto is a postdoc in the Probabilistic Machine Learning Group at Aalto University. His main research interests revolve around the analysis of random objects in complex geometries, where he finds inspiration in optimal transport. He earned his PhD on Geometric Methods in Probabilistic Modelling under the supervision of Åsa Feragen.
• Jean Feydy is a postdoc at the Imperial College London in the team of Michael Bronstein. His main research interests are computational anatomy, optimal transport theory and geometric data analysis. He did his PhD under the direction of Alain Trouvé on Geometric data analysis and developed KeOps, a framework that speeds up geometric computations using symbolic matrices.