An Energy-Based View of Graph Neural Networks

John Young Shin, Prathamesh Dharangutte

Keywords: energy-based models, graph neural networks, stochastic Langevin gradient descent

TL;DR: An energy-based graph neural network is proposed, and its accuracy and robustness is analyzed.

Abstract:

Graph neural networks are a popular variant of neural networks that work with
graph-structured data. In this work, we consider combining graph neural networks
with the energy-based view of Grathwohl et al. (2019) with the aim of obtaining
a more robust classifier. We successfully implement this framework by proposing
a novel method to ensure generation over features as well as the adjacency
matrix and evaluate our method against the standard graph convolutional network
(GCN) architecture (Kipf & Welling (2016)). Our approach obtains comparable
discriminative performance while improving robustness, opening promising new
directions for future research for energy-based graph neural networks.