Variational Inference for Graph Convolutional Networks in the Absence of Graph Data and Adversarial Settings

TL;DR

This paper introduces a variational inference framework that enables graph convolutional networks to operate without explicit graph data and enhances their robustness against adversarial attacks by jointly inferring graph structure and model parameters.

Contribution

It presents a novel probabilistic model and inference algorithm that allows GCNs to learn from data without pre-existing graphs and improves their resilience to adversarial perturbations.

Findings

Outperforms state-of-the-art methods on semi-supervised classification tasks.

Effectively infers graph structure in absence of explicit data.

Increases robustness of GCNs to adversarial attacks.

Abstract

We propose a framework that lifts the capabilities of graph convolutional networks (GCNs) to scenarios where no input graph is given and increases their robustness to adversarial attacks. We formulate a joint probabilistic model that considers a prior distribution over graphs along with a GCN-based likelihood and develop a stochastic variational inference algorithm to estimate the graph posterior and the GCN parameters jointly. To address the problem of propagating gradients through latent variables drawn from discrete distributions, we use their continuous relaxations known as Concrete distributions. We show that, on real datasets, our approach can outperform state-of-the-art Bayesian and non-Bayesian graph neural network algorithms on the task of semi-supervised classification in the absence of graph data and when the network structure is subjected to adversarial perturbations.