# A Comprehensive Survey on Graph Neural Networks

@article{Wu2019ACS, title={A Comprehensive Survey on Graph Neural Networks}, author={Zonghan Wu and Shirui Pan and Fengwen Chen and Guodong Long and Chengqi Zhang and Philip S. Yu}, journal={IEEE Transactions on Neural Networks and Learning Systems}, year={2019}, volume={32}, pages={4-24} }

Deep learning has revolutionized many machine learning tasks in recent years, ranging from image classification and video processing to speech recognition and natural language understanding. The data in these tasks are typically represented in the Euclidean space. However, there is an increasing number of applications, where data are generated from non-Euclidean domains and are represented as graphs with complex relationships and interdependency between objects. The complexity of graph data has… Expand

#### Supplemental Code

Github Repo

Via Papers with Code

Organizing concepts related to machine learning and artificial intelligence

#### Figures, Tables, and Topics from this paper

#### Paper Mentions

#### 1,789 Citations

Graph Neural Networks: Methods, Applications, and Opportunities

- Computer Science
- ArXiv
- 2021

In the last decade or so, we have witnessed deep learning reinvigorating the machine learning field. It has solved many problems in the domains of computer vision, speech recognition, natural… Expand

Graph Neural Networks for Node-Level Predictions

- Computer Science, Mathematics
- ArXiv
- 2020

This work aims to provide an overview of early and modern graph neural network based machine learning methods for node-level prediction tasks and explains the core concepts and detailed explanations for convolutional methods that have had strong impact. Expand

Graph Neural Network and its applications

- Physics
- 2021

In recent years, people have become more interested in expanding deep learning methods on graphs, and a lot of progress has been made in the field. Although traditional deep learning methods have… Expand

Graph Deep Learning: State of the Art and Challenges

- Computer Science
- IEEE Access
- 2021

A review of the state of graph representation learning from the perspective of deep learning and identifies four major challenges in graph deep learning: dynamic and evolving graphs, learning with edge signals and information, graph estimation, and the generalization of graph models. Expand

GraphMDN: Leveraging graph structure and deep learning to solve inverse problems

- Computer Science
- 2021 International Joint Conference on Neural Networks (IJCNN)
- 2021

A Graph Mixture Density Network is developed, which combines graph neural networks with mixture density network outputs and has the advantage of naturally being able to incorporate graph structured information into a neural architecture, as well as the ability to model multi-modal regression targets. Expand

Graph Neural Networks: A Review of Methods and Applications

- Computer Science, Mathematics
- AI Open
- 2020

A detailed review over existing graph neural network models is provided, systematically categorize the applications, and four open problems for future research are proposed. Expand

P3: Distributed Deep Graph Learning at Scale

- Computer Science
- OSDI
- 2021

Graph Neural Networks (GNNs) have gained significant attention in the recent past, and become one of the fastest growing subareas in deep learning. While several new GNN architectures have been… Expand

Dynamic Filters in Graph Convolutional Neural Networks

- Computer Science
- ArXiv
- 2021

A novel method to adapt the behaviour of a GCNN to the input is presented proposing two ways to perform spatial convolution on graphs using input-based filters which are dynamically generated. Expand

Graph convolutional networks: a comprehensive review

- Computer Science
- Computational Social Networks
- 2019

A comprehensive review specifically on the emerging field of graph convolutional networks, which is one of the most prominent graph deep learning models, is conducted and several open challenges are presented and potential directions for future research are discussed. Expand

A Convolutional Neural Network into graph space

- Computer Science
- ArXiv
- 2020

This paper proposes a new convolution neural network architecture, defined directly into graph space, and shows its usability in a back-propagation context and experimental results show that the model performance is at state of the art level on simple tasks. Expand

#### References

SHOWING 1-10 OF 201 REFERENCES

Deep Learning on Graphs: A Survey

- Computer Science, Mathematics
- ArXiv
- 2018

This survey comprehensively review the different types of deep learning methods on graphs by dividing the existing methods into five categories based on their model architectures and training strategies: graph recurrent neural networks, graph convolutional networks,graph autoencoders, graph reinforcement learning, and graph adversarial methods. Expand

Geometric Deep Learning on Graphs and Manifolds Using Mixture Model CNNs

- Computer Science
- 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)
- 2017

This paper proposes a unified framework allowing to generalize CNN architectures to non-Euclidean domains (graphs and manifolds) and learn local, stationary, and compositional task-specific features and test the proposed method on standard tasks from the realms of image-, graph-and 3D shape analysis and show that it consistently outperforms previous approaches. Expand

Multi-dimensional Graph Convolutional Networks

- Computer Science
- SDM
- 2019

This paper proposes a multi-dimensional convolutional neural network model mGCN aiming to capture rich information in learning node-level representations formulti-dimensional graphs and demonstrates the effectiveness of the proposed framework on real-world multi- dimensional graphs. Expand

Hierarchical Graph Representation Learning with Differentiable Pooling

- Computer Science, Mathematics
- NeurIPS
- 2018

DiffPool is proposed, a differentiable graph pooling module that can generate hierarchical representations of graphs and can be combined with various graph neural network architectures in an end-to-end fashion. Expand

An End-to-End Deep Learning Architecture for Graph Classification

- Computer Science
- AAAI
- 2018

This paper designs a localized graph convolution model and shows its connection with two graph kernels, and designs a novel SortPooling layer which sorts graph vertices in a consistent order so that traditional neural networks can be trained on the graphs. Expand

Deep Convolutional Networks on Graph-Structured Data

- Computer Science
- ArXiv
- 2015

This paper develops an extension of Spectral Networks which incorporates a Graph Estimation procedure, that is test on large-scale classification problems, matching or improving over Dropout Networks with far less parameters to estimate. Expand

Large-Scale Learnable Graph Convolutional Networks

- Computer Science, Mathematics
- KDD
- 2018

The proposed LGCL automatically selects a fixed number of neighboring nodes for each feature based on value ranking in order to transform graph data into grid-like structures in 1-D format, thereby enabling the use of regular convolutional operations on generic graphs. Expand

How Powerful are Graph Neural Networks?

- Computer Science, Mathematics
- ICLR
- 2019

This work characterize the discriminative power of popular GNN variants, such as Graph Convolutional Networks and GraphSAGE, and show that they cannot learn to distinguish certain simple graph structures, and develops a simple architecture that is provably the most expressive among the class of GNNs. Expand

Robust Spatial Filtering With Graph Convolutional Neural Networks

- Computer Science
- IEEE Journal of Selected Topics in Signal Processing
- 2017

A novel neural learning framework that is capable of handling both homogeneous and heterogeneous data while retaining the benefits of traditional CNN successes is proposed, which is term Graph-CNNs, which can handle both heterogeneous and homogeneous graph data. Expand

Deeper Insights into Graph Convolutional Networks for Semi-Supervised Learning

- Computer Science, Mathematics
- AAAI
- 2018

It is shown that the graph convolution of the GCN model is actually a special form of Laplacian smoothing, which is the key reason why GCNs work, but it also brings potential concerns of over-smoothing with many convolutional layers. Expand