Graphs for Data Science and Machine Learning
0 likes1,120 views
This document discusses how graphs and graph databases can be used for data science and machine learning. It provides an overview of Neo4j's graph data science capabilities including graph algorithms, machine learning techniques, and real-world use cases. The key points are: 1) Neo4j provides a graph data science library with over 70 graph algorithms and machine learning methods that can be used for tasks like link prediction, node classification, and graph feature engineering. 2) The library allows for both unsupervised and supervised machine learning on graph data in order to identify patterns, anomalies, and make predictions. 3) Real-world examples are presented where companies have used Neo4j's graph data
Graphs for Data Science and Machine Learning
- 1. Neo4j, Inc. All rights reserved 2021 Neo4j, Inc. All rights reserved 2021 1 Graphs for Data Science and Machine Learning Dr. Jim Webber Chief Scientist, Neo4j #neo4j @jimwebber
- 2. Neo4j, Inc. All rights reserved 2021 It’s Not What You Know
- 3. Neo4j, Inc. All rights reserved 2021 It’s Who You Know
- 4. Neo4j, Inc. All rights reserved 2021 It’s Who You Know And Where They Are
- 5. Neo4j, Inc. All rights reserved 2021 5 Higher Pay and More Promotions • People Near Structural Holes • Organizational Misfits Network Structure is Highly Predictive Photo by Helena Lopes on Unsplash “Organizational Misfits and the Origins of Brokerage in Intrafirm Networks” A. Kleinbaum “Structural Holes and Good Ideas” R. Burt
- 6. Neo4j, Inc. All rights reserved 2021 Neo4j, Inc. All rights reserved 2021 6 Relationships are the strongest predictors of behavior But You Can’t Analyse What You Can’t See ● Most data science techniques ignore relationships ● It’s painful to manually engineer connected features from tabular data ● Graphs are built on relationships, so… ● You don’t have to guess at the correlations: with graphs, relationships are built in James Fowler
- 7. Neo4j, Inc. All rights reserved 2021 7 7 Top 10 Tech Trends in Data and Analytics, 16 Feb 2021 According to Garner, “Graphs form the foundation of modern D&A, with capabilities to enhance and improve user collaboration, ML models and explainable AI. The recent Gartner AI in Organizations Survey demonstrates that graph techniques are increasingly prevalent as AI maturity grows, going from 13% adoption when AI maturity is lowest to 48% when maturity is highest.” AI Research Papers Featuring Graph Source: Dimensions Knowledge System 4x Increase in traffic to Neo4j GDS page in 2H- 2020 Analytics & Data Science Interest Exploding in Neo4j Community +4.8m Views on the graph algorithms short video +193k downloads
- 8. Neo4j, Inc. All rights reserved 2021 8 Node Represents an entity in the graph Relationship Connect nodes to each other Property Describes a node or relationship: e.g. name, age, weight etc Wait, what’s a graph? MICA ANDRE Name: “Andre” Born: May 29, 1970 Twitter: “@dan” Name: “Mica” Born: Dec 5, 1975 CAR Brand “Volvo” Model: “V70” Since: Jan 10, 2011 LOVES SISTER BROTHER O W N S D R I V E S
- 9. Neo4j, Inc. All rights reserved 2021 Networks of People Transaction Networks Bought B o u gh t V i e w e d R e t u r n e d Bought Knowledge Networks Pl ay s Lives_in In_sport Likes F a n _ o f Plays_for Risk management, Supply chain, Orders, Payments, etc. Employees, Customers, Suppliers, Partners, Influencers, etc. Enterprise content, Domain specific content, eCommerce content, etc K n o w s Knows Knows K n o w s 9 Everything is Naturally Connected
- 10. Neo4j, Inc. All rights reserved 2021 10 Queries Find the patterns you know exist. Machine Learning Uncover trends and make predictions Visualization Explore, collaborate, and explain Graphs & Data Science Analytics Feature Engineering Data Exploration Graph Data Science Queries Machine Learning Visualization
- 11. Neo4j, Inc. All rights reserved 2021 11 Graphs & Data Science Knowledge Graphs Graph Algorithms Graph Native Machine Learning Find the patterns you’re looking for in connected data Use unsupervised machine learning techniques to identify associations, anomalies, and trends. Use embeddings to learn the features in your graph that you don’t even know are important yet. Train in-graph supervised ML models to predict links, labels, and missing data.
- 12. Neo4j, Inc. All rights reserved 2021 Neo4j’s Graph Data Science Framework Neo4j Graph Data Science Library Neo4j Database Neo4j Bloom Scalable Graph Algorithms & Analytics Workspace Native Graph Creation & Persistence Visual Graph Exploration & Prototyping
- 13. Neo4j, Inc. All rights reserved 2021 ~70 Robust Graph Algorithms & ML methods ● Compute metrics about the topology and connectivity ● Build predictive models to enhance your graph ● Highly parallelized and scale to 10’s of billions of nodes 13 The Neo4j GDS Library Mutable In-Memory Workspace Computational Graph Native Graph Store Efficient & Flexible Analytics Workspace ● Automatically reshapes transactional graphs into an in-memory analytics graph ● Optimized for global traversals and aggregation ● Create workflows and layer algorithms ● Store and manage predictive models in the model catalog
- 14. Neo4j, Inc. All rights reserved 2021 14 ~70 Graph Data Science Techniques in Neo4j Pathfinding & Search • Shortest Path • Single-Source Shortest Path • All Pairs Shortest Path • A* Shortest Path • Yen’s K Shortest Path • Minimum Weight Spanning Tree • K-Spanning Tree (MST) • Random Walk • Breadth & Depth First Search Centrality & Importance • Degree Centrality • Closeness Centrality • Harmonic Centrality • Betweenness Centrality & Approx. • PageRank • Personalized PageRank • ArticleRank • Eigenvector Centrality • Hyperlink Induced Topic Search (HITS) • Influence Maximization (Greedy, CELF) Community Detection • Triangle Count • Local Clustering Coefficient • Connected Components (Union Find) • Strongly Connected Components • Label Propagation • Louvain Modularity • K-1 Coloring • Modularity Optimization • Speaker Listener Label Propagation Supervised Machine Learning • Node Classification • Link Prediction … and more! Heuristic Link Prediction • Adamic Adar • Common Neighbors • Preferential Attachment • Resource Allocations • Same Community • Total Neighbors Similarity • Node Similarity • K-Nearest Neighbors (KNN) • Jaccard Similarity • Cosine Similarity • Pearson Similarity • Euclidean Distance • Approximate Nearest Neighbors (ANN) Graph Embeddings • Node2Vec • FastRP • FastRPExtended • GraphSAGE • Synthetic Graph Generation • Scale Properties • Collapse Paths • One Hot Encoding • Split Relationships • Graph Export • Pregel API (write your own algos)
- 15. Neo4j, Inc. All rights reserved 2021 Our Special Sauce: The Graph Catalogue • Neo4j automates data transformations • Experiment with different data sets, data models • Fast iterations & layering • Production ready features, parallelization & enterprise support • Ability to persist and version data A graph-specific analytics workspace that’s mutable – integrated with a native- graph database Mutable In-Memory Workspace Computational Graph Native Graph Store
- 16. Neo4j, Inc. All rights reserved 2021 Neo4j, Inc. All rights reserved 2021 Right, so how will this improve my machine learning project? 16
- 17. Neo4j, Inc. All rights reserved 2021 Community Detection 17 Neo4j’s Graph Data Science Library Unsupervised Graph Algorithms Clustering Dimension Reduction (generalization) Association Which parts of my graph are connected to each other? Which nodes are most similar? How important is each node? Supervised Machine Learning Node Classification Link Prediction Where will connections form next? What’s the label for this node? Centrality Embeddings Similarity Pathfinding More Algos than any other vendor ONLY in neo4j
- 18. Neo4j, Inc. All rights reserved 2021 Better Predictions with Data You Already Have ● Traditional ML ignores network structure because it’s difficult to extract ● Add graph data to existing ML pipelines to increase accuracy, or ● Graphs use relationships to unlock otherwise unattainable predictions 18 Machine Learning Pipeline
- 19. Neo4j, Inc. All rights reserved 2021 Graphs & Supervised Machine Learning Traditional ML problems where relationships between your data points are important predictive features 19 Predictions influenced by graph structure Predictions about graph structure Enhance your graph by predicting missing data or changes to your graph that will occur in the future
- 20. Neo4j, Inc. All rights reserved 2021 Graphs & Supervised Machine Learning Traditional ML problems where relationships between your data points are important predictive features 20 Predictions influenced by graph structure Predictions about graph structure Enhance your graph by predicting missing data or changes to your graph that will occur in the future
- 21. Neo4j, Inc. All rights reserved 2021 Graphs & Supervised Machine Learning Traditional ML problems where relationships between your data points are important predictive features 21 Predictions influenced by graph structure Predictions about graph structure Enhance your graph by predicting missing data or changes to your graph that will occur in the future
- 22. Neo4j, Inc. All rights reserved 2021 Graphs & Supervised Machine Learning Traditional ML problems where relationships between your data points are important predictive features 22 Predictions influenced by graph structure Predictions about graph structure Enhance your graph by predicting missing data or changes to your graph that will occur in the future
- 23. Neo4j, Inc. All rights reserved 2021 23 Graph Feature Engineering Feature Engineering is how we combine and process data to create new, more meaningful features. Graph algorithms and embeddings translate the connections within your data into the rows and columns you need for ML.
- 24. Neo4j, Inc. All rights reserved 2021 24 In-Graph Machine Learning Node classification: “What kind of node is this?” Link prediction: “Should there be a relationship between these nodes?” Labeled data: Pairs of nodes that are either linked or not Features: Pre-existing attributes, algorithms (pageRank), embedding
- 25. Neo4j, Inc. All rights reserved 2021 25 Node Classification - in Neo4j Load your in- memory graph with labels & features Use nodeClassification.train Specify the property you want to predict and the features for making that prediction Node classification: Predicting a node label or (categorical) property Neo4j Automates the Tricky Parts: 1. Splits data for train & test 2. Builds logistic regression models using the training data & specified parameters to predict the correct label 3. Evaluates the accuracy of the models using the test data 4. Returns the best performing model The predictive model appears in the model catalog, ready to apply to new data
- 26. Neo4j, Inc. All rights reserved 2021 26 Link Prediction - in Neo4j Load your in- memory graph with labels & features Use linkPrediction.train Split your graph into train & test splitRelationships.mutate Link Prediction: Predicting unobserved edges or relationships that will form in the future Neo4j Automates the Tricky Parts: 1. Builds logistic regression models using the training data & specified parameters to predict the correct label 2. Evaluates the accuracy of the models using the test data 3. Returns the best performing model The predictive model appears in the model catalog, ready to apply to new data
- 27. Neo4j, Inc. All rights reserved 2021 27 Machine Learning Models in Neo4j Train a model using your graph and apply it to new or unseen data Not a data model — a predictive model Models live in the Neo4j in the model catalog • Contains versioning information • Input data • Time stamps • Model names • Trained models can be persisted to disk and shared with colleagues ML Models in the Analytics Workspace
- 28. Neo4j, Inc. All rights reserved 2021 28 Neo4j: The Only Completely In-Graph, ML Workflow Graph-Native Feature Engineering Train Predictive Model Queries Algorithms Embeddings 1. Model Type 2. Property Selection 3. Train & Test 4. Model Selection Apply Model to Existing / New Data Use Predictions for Decisions Use Predictions to Enhance the Graph Publish & Share Store Model in Database
- 29. Neo4j, Inc. All rights reserved 2021 29 Neo4j is part of your data ecosystem DATA SOURCES USE CASES INGEST Apache Hop Structured Unstructured DATA ANALYTICS DATA MANAGEMENT Journey Analytics Risk Analytics Churn Analysis What-if Analysis Feature Engineering & ML Fraud Recommendations Data Fabric Data Compliance Data Governance Data Provenance Data Lineage Next Best Case Ontologies Neo4j Bloom Neo4j GDS Library PRODUCT COMPONENTS APOC VISUALIZE AUTO ML DRIVERS & APIs
- 30. Neo4j, Inc. All rights reserved 2021 Neo4j, Inc. All rights reserved 2021 Real World Use Cases 30
- 31. Neo4j, Inc. All rights reserved 2021 31 Graph Data Science Spans Industries and Uses Personalized Recommendations Churn Prediction Market Segmentation Life Sciences Predictive Maintenance Cybersecurity Master Data Management Fraud Detection
- 32. Neo4j, Inc. All rights reserved 2021 32 Accelerate Innovation using Neo4j Graph Data Science From Simple to Highly Sophisticated Data Science Uranus is the third biggest planet Analysis Repeatability Analysis Complexity Full Production Simple, Ad Hoc High Analytics Data Science
- 33. Neo4j, Inc. All rights reserved 2021 33 Accelerate Innovation using Neo4j Graph Data Science From Simple to Highly Sophisticated Data Science Uranus is the third biggest planet R&D: Better health outcomes through machine learning on patient journeys Fraud Detection with graph feature engineering + AutoML Analytics to improve reliability by predicting problems in a supply-chain knowledge graph Analysis Repeatability Analysis Complexity Full Production Simple, Ad Hoc High Analytics Data Science FinServ Customers
- 34. Neo4j, Inc. All rights reserved 2021 34 • Challenge: Difficulty finding faulty components via ad hoc analytics on a vertically integrated supply chain • Solution: Uses a knowledge graph to model and analyze their complex products • Results: ○ Quickly pinpoint root causes of problems ○ Reduced query times from two minutes to seconds ○ Anti-recommendation using graph algorithms to identify and eliminate bad combinations of components Boston Scientific Finding At-Fault Components
- 35. Neo4j, Inc. All rights reserved 2021 35 • Challenge: Graphs are an important predictive signal, but can be challenging to incorporate into production ML • Solution: Use Neo4j for repeatable feature engineering and incorporate results into autoML pipelines. • Results: ○ Identified millions of dollars in previously undetectable fraud ○ Enriched graph with the results of investigations to improve future predictions Top 10 Bank Fighting Fraud
- 36. Neo4j, Inc. All rights reserved 2021 36 AstraZeneca Patient Journey “We used graph algorithms to find patients that had specific journey types and patterns and then find others that are close and similar.” Joseph Roemer Global Commercial IT Insight & Analytics Sr. Director AstraZeneca ● Challenge: How to best intervene sooner for complex diseases that develop over years ● Solution: Neo4j knowledge graph of 3 yrs of visits, tests, & diagnosis with 10’s billions of records. Using graph algorithms and machine learning together. ● Results: ○ Identified journey archetypes and patterns using graph feature engineering as input to ML ○ Revealed journey similarities over time with community detection ○ Found influential touch-points in the journey using graph algorithms
- 37. Neo4j, Inc. All rights reserved 2021 What’s most important and influential in my business? What’s occurring that’s unusual? What’s going to happen next? But traditional approaches to data make it impossible to reveal and effectively use those connections as data sizes become large Predictive signals get lost in big data noise 37 Graph Data Science Answers the BIG Questions Connected Data is Powerful Graph Data Science uses Connections to Answer Critical Questions
- 38. Neo4j, Inc. All rights reserved 2021 38 Neo4j Graph Data Science 70 Graph Algorithms More supported algorithms than any other vendor Graph-Native ML Only commercial offering with full graph ML workflows Humane Experience Automatic transformation from storage to analytics and visualization Scalable Data Science Algorithms running over 10’s billions of nodes in production Extensible Integrate with other data sources and ML platforms Strongest Community 220K+ practioners 72K+ meetups
- 39. Neo4j, Inc. All rights reserved 2021 Neo4j, Inc. All rights reserved 2021 39 Resources Graph Resources ● Video: Advantages of Graph Technology ● Whitepaper: AI & Graph Technology: Enhancing AI with Context & Connections ● Whitepaper: Financial Fraud Detection with Graph Data Science ● Case Study: Meredith Corporation Neo4j Bookshelf ● Graph Databases For Dummies ● Graph Data Science For Dummies ● O’Reilly Graph Algorithms ● O’Reilly Knowledge Graphs
- 40. Neo4j, Inc. All rights reserved 2021 Neo4j, Inc. All rights reserved 2021 40 Thank you! Come see us at Stand B400 Dr. Jim Webber Chief Scientist, Neo4j [email protected]