DeepAI AI Chat
Log In Sign Up

Beltrami Flow and Neural Diffusion on Graphs

10/18/2021
by   Benjamin Paul Chamberlain, et al.
Twitter
74

We propose a novel class of graph neural networks based on the discretised Beltrami flow, a non-Euclidean diffusion PDE. In our model, node features are supplemented with positional encodings derived from the graph topology and jointly evolved by the Beltrami flow, producing simultaneously continuous feature learning and topology evolution. The resulting model generalises many popular graph neural networks and achieves state-of-the-art results on several benchmarks.

READ FULL TEXT

page 1

page 2

page 3

page 4

06/21/2021

GRAND: Graph Neural Diffusion

We present Graph Neural Diffusion (GRAND) that approaches deep learning ...
04/17/2019

Inductive Graph Representation Learning with Recurrent Graph Neural Networks

In this paper, we study the problem of node representation learning with...
05/11/2021

Graph Theory for Metro Traffic Modelling

A unifying graph theoretic framework for the modelling of metro transpor...
04/05/2018

Adaptive Diffusions for Scalable Learning over Graphs

Diffusion-based classifiers such as those relying on the Personalized Pa...
08/12/2020

Quaternion Graph Neural Networks

We consider reducing model parameters and moving beyond the Euclidean sp...
06/02/2022

Invertible Neural Networks for Graph Prediction

In this work, we address conditional generation using deep invertible ne...