Add to my Google Calendar | Learn about Google Calendar

Workshop on Optimal Transport from Theory to Applications -- Interfacing Dynamical Systems, Optimization, and Machine Learning

Optimal transport (OT) is a theory connecting PDEs, geometry, and probability theory. Recent developments
in numerical algorithms for OT problems have further opened new applications, e.g., in
image segmentation, in statistical inference, machine learning, and optimization.


Despite excellent work in various fields trying to bridge the gaps in theory and practice, there is still a significant divide between the researchers working in various mathematical and engineering fields while using the same OT tools. Furthermore, beyond simple static solvers of OT, practitioners in applied fields such as deep learning and control engineering often do not explore the rich and dynamic theory potential of OT. The advantages that OT-based methods bring to applications are also often poorly understood among applied researchers. As a result, the current practice and theory of OT in downstream applications such as machine learning, optimization, and engineering, albeit exciting, remain fragmented and prone to reinventing-the-wheel phenomena.

The OT-DOM workshop remedies this gap by bringing together international experts in OT theory as well as domain experts in applied areas, such as machine learning, optimization, and engineering, in the broader Europe area and beyond, to gather in Berlin, Germany.
The workshop will foster world-level collaboration in the field of OT with potentially impactful downstream applications to deep learning, medical imaging, control and robotics, and other emerging fields.
http://cvgmt.sns.it/event/823/

When
Mon Mar 11 – Fri Mar 15, 2024
Where
Berlin (map)