Bayesian Graph Convolutional Neural Networks Using Non-Parametric Graph Learning

Abstract

Graph convolutional neural networks (GCNN) have been successfully applied to many different graph based learning tasks including node and graph classification, matrix completion, and learning of node embeddings. Despite their impressive performance, the techniques have a limited capability to incorporate the uncertainty in the underlined graph structure. In order to address this issue, a Bayesian GCNN (BGCN) framework was recently proposed. In this framework, the observed graph is considered to be a random realization from a parametric random graph model and the joint Bayesian inference of the graph and GCNN weights is performed. In this paper, we propose a non-parametric generative model for graphs and incorporate it within the BGCN framework. In addition to the observed graph, our approach effectively uses the node features and training labels in the posterior inference of graphs and attains superior or comparable performance in benchmark node classification tasks.

References in corpus (10)

• Deep Convolutional Networks on Graph-Structured Data
• Variational Graph Auto-Encoders
• Probabilistic Backpropagation for Scalable Learning of Bayesian Neural Networks
• Residual Gated Graph ConvNets
• Bayesian Dark Knowledge
• Preconditioned Stochastic Gradient Langevin Dynamics for Deep Neural Networks
• Robust Spatial Filtering with Graph Convolutional Neural Networks
• Dynamic Edge-Conditioned Filters in Convolutional Neural Networks on Graphs
• Dual-Primal Graph Convolutional Networks
• Fisher-Bures Adversary Graph Convolutional Networks

Cited by in corpus (4)

• Bayesian Spatio-Temporal Graph Convolutional Network for Traffic Forecasting
• Graph Posterior Network: Bayesian Predictive Uncertainty for Node Classification
• Bayesian Graph Neural Network for Fast identification of critical nodes in Uncertain Complex Networks
• Graph Signal Processing over a Probability Space of Shift Operators