Gromov Wasserstein

Gromov-Wasserstein (GW) distance is a powerful tool for comparing probability distributions residing in different metric spaces, extending the capabilities of traditional optimal transport methods. Current research focuses on improving the computational efficiency of GW calculations, particularly through dynamic programming and sparsification techniques, and on applying GW to diverse problems such as graph analysis, dimensionality reduction, clustering, and knowledge graph alignment. These advancements are enabling the use of GW in various fields, including computer vision, machine learning, and bioinformatics, where it offers a robust way to compare and analyze complex structured data.

Papers