Optimal Transport for Statistics and Machine Learning
Monday April 4, 2022 | 3.15 pm
Since its introduction more than two centuries ago, optimal transport has flourished into a rich mathematical field allowing us to draw new connections between analysis, geometry, and probability. Recently, thanks to breakneck advances on the computational front, optimal transport has enabled the development of new tools for data analysis, finding applications in a variety of fields ranging from graphics to biology.
Underlying these tools is a new machine learning paradigm where the goal is integrate multiple data sources. This talk will illustrate this new paradigm in light of several applications and discuss some of the statistical challenges associated with it.
Philippe Rigollet is a Professor of Mathematics at MIT, where he is also a member of the Statistics and Data Science Center, a Principal Investigator at the Laboratory for Information and Decision Systems. He is also an affiliate member of the Broad Institute. He was previously an Assistant Professor at Princeton and a Postdoc at Georgia Tech. He is currently on leave from MIT and holds a Chair from the Fondation des Sciences Mathématiques de Paris.
Philippe’s academic interests are on the mathematical aspects of data and his work aims at designing methods to process complex data using geometric ideas. His contributions have been recognized by a CAREER award from the National Science Foundation, a Best Paper Award at the Conference on Learning Theory, and a Medallion Lecture at the Joint Statistical Meetings. He is an elected Fellow of the Institute for Mathematical Statistics.