Constant Curvature Graph Convolutional Networks

11/12/2019
by   Gregor Bachmann, et al.
0

Interest has been rising lately towards methods representing data in non-Euclidean spaces, e.g. hyperbolic or spherical, that provide specific inductive biases useful for certain real-world data properties, e.g. scale-free, hierarchical or cyclical. However, the popular graph neural networks are currently limited in modeling data only via Euclidean geometry and associated vector space operations. Here, we bridge this gap by proposing mathematically grounded generalizations of graph convolutional networks (GCN) to (products of) constant curvature spaces. We do this by i) introducing a unified formalism that can interpolate smoothly between all geometries of constant curvature, ii) leveraging gyro-barycentric coordinates that generalize the classic Euclidean concept of the center of mass. Our class of models smoothly recover their Euclidean counterparts when the curvature goes to zero from either side. Empirically, we outperform Euclidean GCNs in the tasks of node classification and distortion minimization for symbolic data exhibiting non-Euclidean behavior, according to their discrete curvature.

READ FULL TEXT

page 1

page 2

page 3

page 4

research
10/28/2019

Hyperbolic Graph Convolutional Neural Networks

Graph convolutional neural networks (GCNs) embed nodes in a graph into E...
research
09/08/2023

Curve Your Attention: Mixed-Curvature Transformers for Graph Representation Learning

Real-world graphs naturally exhibit hierarchical or cyclical structures ...
research
08/05/2019

Rendering Non-Euclidean Geometry in Real-Time Using Spherical and Hyperbolic Trigonometry

This paper introduces a method of calculating and rendering shapes in a ...
research
06/06/2021

Semi-Riemannian Graph Convolutional Networks

Graph Convolutional Networks (GCNs) are typically studied through the le...
research
02/19/2021

Linear Classifiers in Mixed Constant Curvature Spaces

Embedding methods for mixed-curvature spaces are powerful techniques for...
research
12/02/2022

A Cosine Rule-Based Discrete Sectional Curvature for Graphs

How does one generalize differential geometric constructs such as curvat...
research
06/23/2023

Curvature-enhanced Graph Convolutional Network for Biomolecular Interaction Prediction

Geometric deep learning has demonstrated a great potential in non-Euclid...

Please sign up or login with your details

Forgot password? Click here to reset