Ollivier-Ricci Curvature of Riemannian Manifolds and Directed Graphs with Applications to Graph Neural Networks
The episode discusses a paper introducing Ollivier-Ricci curvature, an extension of classical Ricci curvature to general metric spaces like graphs, using one-Wasserstein distance for comparison. The hosts detail how this concept is applied to undirected and directed graphs, focusing on its use in guiding Graph Neural Networks for tasks like community detection and network analysis.