
A novel method for representing graph-structured data in low-dimensional metric spaces that combines the efficiency of optimization-based embeddings and the expressiveness of neural network-based approximations to accurately reflect the topological distances within the graph.


coremsg

preserving-topological-distances-in-low-dimensional-graph-vertex-embeddings-through-regularization-and-flexible-distance-functions


Preserving Topological Distances in Low-Dimensional Graph Vertex Embeddings through Regularization and Flexible Distance Functions
