Deep learning on graphs / Yao Ma, Jiliang Tang.

Format:

Book

Author/Creator:

Ma, Yao, author.

Tang, Jiliang, author.

Language:

English

Subjects (All):

Machine learning.

Graph algorithms.

Physical Description:

1 online resource (xviii, 320 pages) : digital, PDF file(s).

Edition:

1st ed.

Place of Publication:

Cambridge : Cambridge University Press, 2021.

Summary:

Deep learning on graphs has become one of the hottest topics in machine learning. The book consists of four parts to best accommodate our readers with diverse backgrounds and purposes of reading. Part 1 introduces basic concepts of graphs and deep learning; Part 2 discusses the most established methods from the basic to advanced settings; Part 3 presents the most typical applications including natural language processing, computer vision, data mining, biochemistry and healthcare; and Part 4 describes advances of methods and applications that tend to be important and promising for future research. The book is self-contained, making it accessible to a broader range of readers including (1) senior undergraduate and graduate students; (2) practitioners and project managers who want to adopt graph neural networks into their products and platforms; and (3) researchers without a computer science background who want to use graph neural networks to advance their disciplines.

Contents:

Cover

Half-title

Title page

Copyright information

Contents

Preface

Acknowledgments

1 Deep Learning on Graphs: An Introduction

1.1 Introduction

1.2 Why Deep Learning on Graphs?

1.3 What Content Is Covered?

1.4 Who Should Read This Book?

1.5 Feature Learning on Graphs: A Brief History

1.5.1 Feature Selection on Graphs

1.5.2 Representation Learning on Graphs

1.6 Conclusion

1.7 Further Reading

Part I Foundations

2 Foundations of Graphs

2.1 Introduction

2.2 Graph Representations

2.3 Properties and Measures

2.3.1 Degree

2.3.2 Connectivity

2.3.3 Centrality

2.4 Spectral Graph Theory

2.4.1 Laplacian Matrix

2.4.2 The Eigenvalues and Eigenvectors of the Laplacian Matrix

2.5 Graph Signal Processing

2.5.1 Graph Fourier Transform

2.6 Complex Graphs

2.6.1 Heterogeneous Graphs

2.6.2 Bipartite Graphs

2.6.3 Multidimensional Graphs

2.6.4 Signed Graphs

2.6.5 Hypergraphs

2.6.6 Dynamic Graphs

2.7 Computational Tasks on Graphs

2.7.1 Node-Focused Tasks

2.7.2 Graph-Focused Tasks

2.8 Conclusion

2.9 Further Reading

3 Foundations of Deep Learning

3.1 Introduction

3.2 Deep Feedforward Networks

3.2.1 The Architecture

3.2.2 Activation Functions

3.2.3 Output Layer and Loss Function

3.3 Convolutional Neural Networks

3.3.1 The Convolution Operation and Convolutional Layer

3.3.2 Convolutional Layers in Practice

3.3.3 Nonlinear Activation Layer

3.3.4 Pooling Layer

3.3.5 An Overall CNN Framework

3.4 Recurrent Neural Networks

3.4.1 The Architecture of Traditional RNNs

3.4.2 Long Short-Term Memory

3.4.3 Gated Recurrent Unit

3.5 Autoencoders

3.5.1 Undercomplete Autoencoders

3.5.2 Regularized Autoencoders

3.6 Training Deep Neural Networks

3.6.1 Training with Gradient Descent

3.6.2 Backpropagation.

3.6.3 Preventing Overfitting

3.7 Conclusion

3.8 Further Reading

Part II Methods

4 Graph Embedding

4.1 Introduction

4.2 Graph Embedding for Simple Graphs

4.2.1 Preserving Node Co-occurrence

4.2.2 Preserving Structural Role

4.2.3 Preserving Node Status

4.2.4 Preserving Community Structure

4.3 Graph Embedding on Complex Graphs

4.3.1 Heterogeneous Graph Embedding

4.3.2 Bipartite Graph Embedding

4.3.3 Multidimensional Graph Embedding

4.3.4 Signed Graph Embedding

4.3.5 Hypergraph Embedding

4.3.6 Dynamic Graph Embedding

4.4 Conclusion

4.5 Further Reading

5 Graph Neural Networks

5.1 Introduction

5.2 The General GNN Frameworks

5.2.1 A General Framework for Node-Focused Tasks

5.2.2 A General Framework for Graph-Focused Tasks

5.3 Graph Filters

5.3.1 Spectral-Based Graph Filters

5.3.2 Spatial-Based Graph Filters

5.4 Graph Pooling

5.4.1 Flat Graph Pooling

5.4.2 Hierarchical Graph Pooling

5.5 Parameter Learning for Graph Neural Networks

5.5.1 Parameter Learning for Node Classification

5.5.2 Parameter Learning for Graph Classification

5.6 Conclusion

5.7 Further Reading

6 Robust Graph Neural Networks

6.1 Introduction

6.2 Graph Adversarial Attacks

6.2.1 Taxonomy of Graph Adversarial Attacks

6.2.2 White-Box Attack

6.2.3 Gray-Box Attack

6.2.4 Black-Box Attack

6.3 Graph Adversarial Defenses

6.3.1 Graph Adversarial Training

6.3.2 Graph Purification

6.3.3 Graph Attention

6.3.4 Graph Structure Learning

6.4 Conclusion

6.5 Further Reading

7 Scalable Graph Neural Networks

7.1 Introduction

7.2 Node-wise Sampling Methods

7.3 Layer-wise Sampling Methods

7.4 Subgraph-wise Sampling Methods

7.5 Conclusion

7.6 Further Reading

8 Graph Neural Networks for Complex Graphs

8.1 Introduction.

8.2 Heterogeneous Graph Neural Networks

8.3 Bipartite Graph Neural Networks

8.4 Multidimensional Graph Neural Networks

8.5 Signed Graph Neural Networks

8.6 Hypergraph Neural Networks

8.7 Dynamic Graph Neural Networks

8.8 Conclusion

8.9 Further Reading

9 Beyond GNNs: More Deep Models on Graphs

9.1 Introduction

9.2 Autoencoders on Graphs

9.3 Recurrent Neural Networks on Graphs

9.4 Variational Autoencoders on Graphs

9.4.1 Variational Autoencoders for Node Representation Learning

9.4.2 Variational Autoencoders for Graph Generation

9.5 Generative Adversarial Networks on Graphs

9.5.1 Generative Adversarial Networks for Node Representation Learning

9.5.2 Generative Adversarial Networks for Graph Generation

9.6 Conclusion

9.7 Further Reading

Part III Applications

10 Graph Neural Networks in Natural Language Processing

10.1 Introduction

10.2 Semantic Role Labeling

10.3 Neural Machine Translation

10.4 Relation Extraction

10.5 Question Answering

10.5.1 The Multihop QA Task

10.5.2 Entity-GCN

10.6 Graph to Sequence Learning

10.7 Graph Neural Networks on Knowledge Graphs

10.7.1 Graph Filters for Knowledge Graphs

10.7.2 Transforming Knowledge Graphs to Simple Graphs

10.7.3 Knowledge Graph Completion

10.8 Conclusion

10.9 Further Reading

11 Graph Neural Networks in Computer Vision

11.1 Introduction

11.2 Visual Question Answering

11.2.1 Images as Graphs

11.2.2 Images and Questions as Graphs

11.3 Skeleton-Based Action Recognition

11.4 Image Classification

11.4.1 Zero-Shot Image Classification

11.4.2 Few-Shot Image Classification

11.4.3 Multilabel Image Classification

11.5 Point Cloud Learning

11.6 Conclusion

11.7 Further Reading

12 Graph Neural Networks in Data Mining

12.1 Introduction

12.2 Web Data Mining.

12.2.1 Social Network Analysis

12.2.2 Recommender Systems

12.3 Urban Data Mining

12.3.1 Traffic Prediction

12.3.2 Air Quality Forecasting

12.4 Cybersecurity Data Mining

12.4.1 Malicious Account Detection

12.4.2 Fake News Detection

12.5 Conclusion

12.6 Further Reading

13 Graph Neural Networks in Biochemistry and Health Care

13.1 Introduction

13.2 Drug Development and Discovery

13.2.1 Molecule Representation Learning

13.2.2 Protein Interface Prediction

13.2.3 Drug-Target Binding Affinity Prediction

13.3 Drug Similarity Integration

13.4 Polypharmacy Side Effect Prediction

13.5 Disease Prediction

13.6 Conclusion

13.7 Further Reading

Part IV Advances

14 Advanced Topics in Graph Neural Networks

14.1 Introduction

14.2 Deeper Graph Neural Networks

14.2.1 Jumping Knowledge

14.2.2 DropEdge

14.2.3 PairNorm

14.3 Exploring Unlabeled Data via Self-Supervised Learning

14.3.1 Node-Focused Tasks

14.3.2 Graph-Focused Tasks

14.4 Expressiveness of Graph Neural Networks

14.4.1 Weisfeiler-Lehman Test

14.4.2 Expressiveness

14.5 Conclusion

14.6 Further Reading

15 Advanced Applications in Graph Neural Networks

15.1 Introduction

15.2 Combinatorial Optimization on Graphs

15.3 Learning Program Representations

15.4 Reasoning Interacting Dynamical Systems in Physics

15.5 Conclusion

15.6 Further Reading

Bibliography

Index.

Notes:

Title from publisher's bibliographic system (viewed on 07 Oct 2021).

Includes bibliographical references and index.

Other Format:

Print version:

ISBN:

1-108-93482-X

1-108-92418-2

OCLC:

1267537609