Optimal transport: applications and statistical modelling

Optimal transport compares and analyses probability distributions while respecting the geometry of the underlying space, which makes its distances interpretable and structure-aware. It underpins methods in generative modelling, domain adaptation, clustering and distributional robustness, and enables principled interpolation, alignment and generation of datasets in machine learning, imaging and scientific data analysis. These strengths come with substantial computational costs in high-dimensional and large-scale settings. The workshop gathers recent advances in the applications and statistical aspects of optimal transport, emphasising structures and challenges shared across domains.