GMNN: Graph Markov Neural Networks

Abstract

This paper studies semi-supervised object classification in relational data, which is a fundamental problem in relational data modeling. The problem has been extensively studied in the literature of both statistical relational learning (e.g. relational Markov networks) and graph neural networks (e.g. graph convolutional networks). Statistical relational learning methods can effectively model the dependency of object labels through conditional random fields for collective classification, whereas graph neural networks learn effective object representations for classification through end-to-end training. In this paper, we propose the Graph Markov Neural Network (GMNN) that combines the advantages of both worlds. A GMNN models the joint distribution of object labels with a conditional random field, which can be effectively trained with the variational EM algorithm. In the E-step, one graph neural network learns effective object representations for approximating the posterior distributions of object labels. In the M-step, another graph neural network is used to model the local label dependency. Experiments on object classification, link classification, and unsupervised node representation learning show that GMNN achieves state-of-the-art results.

arXiv commenticml 2019

Cited by in corpus (53)

• Geometric Deep Learning: Grids, Groups, Graphs, Geodesics, and Gauges
• A Gentle Introduction to Deep Learning for Graphs
• PairNorm: Tackling Oversmoothing in GNNs
• Beyond Homophily in Graph Neural Networks: Current Limitations and Effective Designs
• Graph Neural Networks: Taxonomy, Advances and Trends
• Combining Label Propagation and Simple Models Out-performs Graph Neural Networks
• Unifying Graph Convolutional Neural Networks and Label Propagation
• Probabilistic Logic Neural Networks for Reasoning
• A Note on Over-Smoothing for Graph Neural Networks
• RNNLogic: Learning Logic Rules for Reasoning on Knowledge Graphs
• Residual Correlation in Graph Neural Network Regression
• Few-shot Relation Extraction via Bayesian Meta-learning on Relation Graphs
• GraphAIR: Graph Representation Learning with Neighborhood Aggregation and Interaction
• A Survey of Community Detection Approaches: From Statistical Modeling to Deep Learning
• Comparison of biomedical relationship extraction methods and models for knowledge graph creation
• Graph Convolutional Neural Networks with Node Transition Probability-based Message Passing and DropNode Regularization
• Continuous Graph Neural Networks
• Efficient Probabilistic Logic Reasoning with Graph Neural Networks
• From Machine Learning to Robotics: Challenges and Opportunities for Embodied Intelligence
• Adaptive Dependency Learning Graph Neural Networks
• Bayesian graph convolutional neural networks via tempered MCMC
• Sub-graph Contrast for Scalable Self-Supervised Graph Representation Learning
• Revisiting Bayesian Autoencoders with MCMC
• Group-Wise Semantic Mining for Weakly Supervised Semantic Segmentation
• Predicting cognitive scores with graph neural networks through sample selection learning
• Can Graph Neural Networks Help Logic Reasoning?
• Extract the Knowledge of Graph Neural Networks and Go Beyond it: An Effective Knowledge Distillation Framework
• GENN: Predicting Correlated Drug-drug Interactions with Graph Energy Neural Networks
• Conditional Temporal Neural Processes with Covariance Loss
• Calibrating and Improving Graph Contrastive Learning
• Joint Learning of Graph Representation and Node Features in Graph Convolutional Neural Networks
• Self-supervised Graph Representation Learning via Bootstrapping
• Markov-Lipschitz Deep Learning
• VigDet: Knowledge Informed Neural Temporal Point Process for Coordination Detection on Social Media
• HighwayGraph: Modelling Long-distance Node Relations for Improving General Graph Neural Network
• Semi-Supervised Node Classification on Graphs: Markov Random Fields vs. Graph Neural Networks
• Graph Neural Networks with Feature and Structure Aware Random Walk
• Learning Attribute-Structure Co-Evolutions in Dynamic Graphs
• Multi-task Self-distillation for Graph-based Semi-Supervised Learning
• Structure-Aware Label Smoothing for Graph Neural Networks
• CopulaGNN: Towards Integrating Representational and Correlational Roles of Graphs in Graph Neural Networks
• Mutual Information Maximization in Graph Neural Networks
• Region-based Energy Neural Network for Approximate Inference
• Multiscale Laplacian Learning
• Adversarially-learned Inference via an Ensemble of Discrete Undirected Graphical Models
• A Collective Learning Framework to Boost GNN Expressiveness
• Barlow Graph Auto-Encoder for Unsupervised Network Embedding
• Interferometric Graph Transform for Community Labeling
• Modeling Graph Node Correlations with Neighbor Mixture Models
• Label-Consistency based Graph Neural Networks for Semi-supervised Node Classification
• LFGCN: Levitating over Graphs with Levy Flights
• Community-preserving Graph Convolutions for Structural and Functional Joint Embedding of Brain Networks
• Estimating Aggregate Properties In Relational Networks With Unobserved Data