Symmetric Spaces for Graph Embeddings: A Finsler-Riemannian Approach

       Federico Lopez 1 Beatrice Pozzetti 2 Steve Trettel 3 Michael Strube 1 Anna Wienhard 2

             Abstract
   Learning faithful graph representations as sets of
   vertex embeddings has become a fundamental in-
   termediary step in a wide range of machine learn-
   ing applications. We propose the systematic use
   of symmetric spaces in representation learning, a
   class encompassing many of the previously used
   ...