book

Applied Deep Learning on Graphs

Name: Applied Deep Learning on Graphs
ISBN: 9781835885963

by Lakshya Khandelwal, Subhajoy Das

December 2024

Intermediate to advanced

250 pages

6h 57m

English

Packt Publishing

Read now

Unlock full access

Applied Deep Learning on Graphs
Contributors
About the authors
About the reviewers
Preface
Who this book is forWhat this book coversTo get the most out of this bookDownload the example code filesConventions usedGet in touchShare Your Thoughts
Part 1: Foundations of Graph Learning
Chapter 1: Introduction to Graph Learning
Do we need graphs?A case studyFormalizing graphsDefinition and semanticsTypes and properties of graphsDirected graphsBipartite graphsConnected graphsWeighted graphsSubgraphsCentralityCommunity structureIsomorphismGraph data structuresAdjacency matrixAdjacency listTraditional graph-based solutionsSearchingPartitioningPath optimizationThe need for representation learningGNNs and the need for a separate verticalSummary
Chapter 2: Graph Learning in the Real World
Node-level learningNode classificationNode regressionNode clusteringNode anomaly detectionEdge-level learningEdge classificationEdge regressionEdge clusteringEdge outlier detectionGraph-level learningGraph-level representationsReal-world applicationsRecommender systemsKnowledge graphsEntity and relationship embeddingsSemantic similarity and entity resolutionSome other applicationsSummary
Chapter 3: Graph Representation Learning
Representation learning – what is it?Graph representation learningA framework for graph learningDeepWalkRandom walk – the what and the whyEstimating the node embeddingsNode2VecGraph traversal approachesFinalizing the random walk strategyNode2Vec versus DeepWalkLimitations of shallow encodingsSummary
Part 2: Advanced Graph Learning Techniques
Chapter 4: Deep Learning Models for Graphs
Technical requirementsMessage passing in graphsDecoding GNNsGCNsUsing GCNs for different graph tasksGraphSAGEGATsAttention networksAttention coefficients computationAggregation of neighbor featuresMulti-head attentionStacking GAT layersSummary

Chapter 5: Graph Deep Learning Challenges
Data-related challengesHeterogeneity in graph structuresDynamic and evolving graphsNoisy and incomplete graph dataModel architecture challengesCapturing long-range dependenciesDepth limitation in GNNsOver-smoothing and over-squashingBalancing local and global informationFacing a model architecture challenge – an exampleComputational challengesScalability issues for large graphsMemory constraints in graph processingParallel and distributed computing for graphsTask-specific challengesNode classification in imbalanced graphsLink prediction in sparse graphsGraph generation and reconstructionGraph matching and alignmentInterpretability and explainabilityExplaining GNN decisionsVisualizing graph embeddingsSummary
Chapter 6: Harnessing Large Language Models for Graph Learning
Understanding LLMsTextual data in graphsLeveraging InstructGLMLLMs for graph learningLLMs as enhancersLLMs as predictorsIntegrating RAG with graph learningAdvantages of graph RAG (GRAG) approachesChallenges in integrating LLMs with graph learningSummary
Part 3: Practical Applications and Implementation
Chapter 7: Graph Deep Learning in Practice
Setting up the environmentCreating the graph datasetNode classification – predicting student interestsLink prediction – recommending new friendshipsSummary
Chapter 8: Graph Deep Learning for Natural Language Processing
Graph structures in NLPImportance of graph representations in languageTypes of linguistic graphsGraph-based text summarizationExtractive summarization using graph centralityAbstractive summarization with graph-to-sequence modelsInformation extraction (IE) using GNNsEvent extractionOpen IEAdvantages of GNN-based IEGraph-based dialogue systemsDialogue state tracking with GNNsGraph-enhanced response generationKnowledge-grounded conversations using graphsGraph-based dialogue policy learningSummary
Chapter 9: Building Recommendation Systems Using Graph Deep Learning
Fundamentals of recommendation systemsTypes of recommendation systemsKey metrics and evaluationGraph structures in recommendation systemsUser-item interaction graphsIncorporating side informationTemporal graphsMulti-relational graphsGraph-based recommendation modelsMF with graph regularizationGraph neural network modelsTraining graph deep learning modelsData preprocessingModel training techniquesScalability and optimizationMini-batch training with neighborhood samplingDistributed trainingExplainability in graph-based recommendationsAttention mechanisms for interpretabilityPath-based explanationsThe cold start problemGraph embedding transferContent-based feature integrationSummary
Chapter 10: Graph Deep Learning for Computer Vision
Traditional CV approaches versus graph-based approachesGraph construction for visual dataPixel-level graphsSuperpixel-based graphsObject-level graphsScene graphsComparing different graph construction methodsGNNs for image classificationGraph convolutional networks for image dataAttention mechanisms in graph-based image classificationHierarchical graph representations for multi-scale feature learningGNNs versus CNNsObject detection and segmentation using GNNsGraph-based object proposal generationInstance segmentation with GNNsPanoptic segmentation using graph-structured outputsMulti-modal learning with GNNsIntegrating visual and textual information using graphsCross-modal retrieval using graph-based representationsGNNs for visual-language navigationLimitations and next stepsScalability issues in large-scale visual datasetsEfficient graph construction and updating for real-time applicationsIntegrating graph-based methods with other deep learning approachesNew applications and research opportunitiesSummary
Part 4: Future Directions
Chapter 11: Emerging Applications
Biology and healthcareProtein-protein interaction networksDrug discovery and developmentDisease prediction and progression modelingBrain connectomics analysisGenomics and gene regulatory networksSocial network analysisCommunity detectionInfluence propagation modelingUser behavior predictionFake news detectionFinancial servicesFraud detection in transaction networksCredit risk assessmentStock market prediction using company relationship graphsAnti-money laundering systemsPersonalized financial recommendationsSystemic risk assessmentCybersecurityWhy graphs for cybersecurity?Network intrusion detectionMalware detectionEnergy systemsGraph representation of energy systemsLoad forecastingFault detection and localizationOptimal power flowRenewable energy forecastingEnergy storage managementVulnerability assessmentIoTDevice interaction modelingAnomaly detection in sensor networksPredictive maintenanceSmart home applicationsLegal governance and complianceKnowledge graph construction for legal and regulatory dataAutomated compliance monitoring and risk assessmentLegal document analysis and contract managementRegulatory intelligence and policy impact assessmentSummary
Chapter 12: The Future of Graph Learning
Emerging trends and directionsScalability and efficiencyInterpretability and explainabilityDynamic and temporal graphsHeterogeneous and multi-modal graphsAdvanced architectures and techniquesGraph transformers and attention mechanismsGraph generative modelsFew-shot and zero-shot learning on graphsReinforcement learning on graphsIntegration with other AI domainsGraph learning and LLMsFederated graph learningQuantum GNNsPotential breakthroughs and long-term visionArtificial general intelligence and graphsNeuromorphic computing with graphsGraph learning in the MetaverseInterdisciplinary applicationsSummary
Index
Why subscribe?
Other Books You May EnjoyPackt is searching for authors like youShare Your ThoughtsDownload a free PDF copy of this book

Overview

Dive into the increasing role of graphs in deep learning with 'Applied Deep Learning on Graphs'. The book provides a structured learning journey from fundamental graph concepts to cutting-edge graph neural networks, ideal for tackling complex interconnected data. With real-world examples and implementation of modern algorithms, you will master the use of graphs for transformative applications.

What this Book will help me do

Understand and utilize graph-centric approaches to solve various problems.
Learn modern techniques for graph representation learning and neural network architectures.
Explore practical applications of graph deep learning in areas like NLP and recommendation systems.
Gain insight into the challenges of scalability and interpretability in graph deep learning.
Develop skills to implement and optimize graph-based models in real-world scenarios using PyTorch Geometric.

Author(s)

Lakshya Khandelwal and Subhajoy Das bring invaluable experience from their tenure at top-tech firms including Walmart, Adobe, and Samsung. Experts in applied graph deep learning, their extensive knowledge in machine learning, coupled with a talent for clear explanation, makes this book a powerful resource for practitioners of AI and data science.

Who is it for?

This book is written for data scientists and machine learning practitioners looking to enhance their skillset in graph-based data analysis. If you are a researcher exploring graphs or a software engineer developing graph applications, this book is invaluable. A fundamental understanding of machine learning and Python is required, making it suitable for intermediate-level learners. Expert guidance will help you achieve your goals.

Become an O’Reilly member and get unlimited access to this title plus top books and audiobooks from O’Reilly and nearly 200 top publishers, thousands of courses curated by job role, 150+ live events each month,
and much more.

Read now

Unlock full access

More than 5,000 organizations count on O’Reilly

O’Reilly covers everything we've got, with content to help us build a world-class technology community, upgrade the capabilities and competencies of our teams, and improve overall team performance as well as their engagement.

Julian F.

Head of Cybersecurity

I wanted to learn C and C++, but it didn't click for me until I picked up an O'Reilly book. When I went on the O’Reilly platform, I was astonished to find all the books there, plus live events and sandboxes so you could play around with the technology.

Addison B.

Field Engineer

I’ve been on the O’Reilly platform for more than eight years. I use a couple of learning platforms, but I'm on O'Reilly more than anybody else. When you're there, you start learning. I'm never disappointed.

Amir M.

Data Platform Tech Lead

I'm always learning. So when I got on to O'Reilly, I was like a kid in a candy store. There are playlists. There are answers. There's on-demand training. It's worth its weight in gold, in terms of what it allows me to do.

Mark W.

Embedded Software Engineer

Publisher Resources

ISBN: 9781835885963

Cloud Computing

Data Engineering

Data Science

AI & ML

Programming Languages

Software Architecture

IT/Ops

Security

Design

Business

Soft Skills