Muutke küpsiste eelistusi

Graph-Powered Machine Learning [Pehme köide]

2.67/5 (12 hinnangut Goodreads-ist)
  • Formaat: Paperback / softback, 503 pages, kõrgus x laius x paksus: 236x185x30 mm, kaal: 860 g
  • Ilmumisaeg: 15-Nov-2021
  • Kirjastus: Manning Publications
  • ISBN-10: 1617295647
  • ISBN-13: 9781617295645
Teised raamatud teemal:
Graph-Powered Machine LearningSuurem pilt
  • Formaat: Paperback / softback, 503 pages, kõrgus x laius x paksus: 236x185x30 mm, kaal: 860 g
  • Ilmumisaeg: 15-Nov-2021
  • Kirjastus: Manning Publications
  • ISBN-10: 1617295647
  • ISBN-13: 9781617295645
Teised raamatud teemal:
Püsilink: https://www.kriso.ee/db/9781617295645.html
Märksõnad:
At its core, machine learning is about efficiently identifying patterns and relationships in data. Many tasks, such as finding associations among terms so you can make accurate search recommendations or locating individuals within a social network who have similar interests, are naturally expressed as graphs.

 

Graph-Powered Machine Learning introduces you to graph technology concepts, highlighting the role of graphs in machine learning and big data platforms. Youll get an in-depth look at techniques including data source modeling, algorithm design, link analysis, classification, and clustering. As you master the core concepts, youll explore three end-to-end projects that illustrate architectures, best design practices,

optimization approaches, and common pitfalls.

 

Key Features

·   The lifecycle of a machine learning project

·   Three end-to-end applications

·   Graphs in big data platforms

·   Data source modeling

·   Natural language processing, recommendations, and relevant search

·   Optimization methods

 

Readers comfortable with machine learning basics.

 

About the technology

By organizing and analyzing your data as graphs, your applications work more fluidly with graph-centric algorithms like nearest neighbor or page rank where its important to quickly identify and exploit relevant relationships. Modern graph data stores, like Neo4j or Amazon Neptune, are readily available tools that support graph-powered machine learning.

 

Alessandro Negro is a Chief Scientist at GraphAware. With extensive experience in software development, software architecture, and data management, he has been a speaker at many conferences, such as Java One, Oracle Open World, and Graph Connect. He holds a Ph.D. in Computer Science and has authored several publications on graph-based machine learning.
Foreword xiii
Preface xv
Acknowledgments xvii
About this Book xix
About the Author xxiii
About the Cover Illustration xxiv
PART 1 INTRODUCTION
1(262)
1 Machine learning and graphs: An introduction
3(27)
1.1 Machine learning project life cycle
5(5)
Business understanding
7(1)
Data understanding
8(1)
Data preparation
8(1)
Modeling
9(1)
Evaluation
9(1)
Deployment
9(1)
1.2 Machine learning challenges
10(5)
The source of truth
10(3)
Performance
13(1)
Storing the model
14(1)
Realtime
14(1)
1.3 Graphs
15(8)
What is a graph?
15(2)
Graphs as models of networks
17(6)
1.4 The role of graphs in machine learning
23(4)
Data management
25(1)
Data analysis
25(1)
Data visualization
26(1)
1.5 Book mental model
27(3)
2 Graph data engineering
30(41)
2.1 Working with big data
33(7)
Volume
34(2)
Velocity
36(2)
Variety
38(1)
Veracity
39(1)
2.2 Graphs in the big data platform
40(13)
Graphs are valuable for big data
41(7)
Graphs are valuable for master data management
48(5)
2.3 Graph databases
53(18)
Graph database management
54(3)
Sharding
57(3)
Replication
60(1)
Native vs. non-native graph databases
61(6)
Label property graphs
67(4)
3 Graphs in machine learning applications
71(42)
3.1 Graphs in the machine learning workflow
73(3)
3.2 Managing data sources
76(17)
Monitor a subject
79(3)
Detect a fraud
82(3)
Identify risks in a supply chain
85(2)
Recommend items
87(6)
3.3 Algorithms
93(7)
Identify risks in a supply chain
93(3)
Find keywords in a document
96(2)
Monitor a subject
98(2)
3.4 Storing and accessing machine learning models
100(6)
Recommend items
101(2)
Monitoring a subject
103(3)
3.5 Visualization
106(3)
3.6 Leftover: Deep learning and graph neural networks
109(4)
PART 2 RECOMMENDATIONS
113(6)
4 Content-based recommendations
119(47)
4.1 Representing item features
122(14)
4.2 User modeling
136(7)
4.3 Providing recommendations
143(21)
4.4 Advantages of the graph approach
164(2)
5 Collaborative filtering
166(36)
5.1 Collaborative filtering recommendations
170(2)
5.2 Creating the bipartite graph for the User-Item dataset
172(5)
5.3 Computing the nearest neighbor network
177(12)
5.4 Providing recommendations
189(5)
5.5 Dealing with the cold-start problem
194(4)
5.6 Advantages of the graph approach
198(4)
6 Session-based recommendations
202(25)
6.1 The session-based approach
203(3)
6.2 The events chain and the session graph
206(6)
6.3 Providing recommendations
212(12)
Item-based k-NN
213(6)
Session-based k-NN
219(5)
6.4 Advantages of the graph approach
224(3)
7 Context-aware and hybrid recommendations
227(36)
7.1 The context-based approach
228(26)
Representing contextual information
231(4)
Providing recommendations
235(18)
Advantages of the graph approach
253(1)
7.2 Hybrid recommendation engines
254(9)
Multiple models, single graph
256(2)
Providing recommendations
258(2)
Advantages of the graph approach
260(3)
PART 3 FIGHTING FRAUD
263(94)
8 Basic approaches to graphpowered fraud detection
265(30)
8.1 Fraud prevention and detection
267(4)
8.2 The role of graphs in fighting fraud
271(8)
8.3 Warm-up: Basic approaches
279(16)
Finding the origin point of credit card fraud
279(8)
Identifying a fraud ring
287(6)
Advantages of the graph approach
293(2)
9 Proximity-based algorithms
295(25)
9.1 Proximity-based algorithms: An introduction
296(2)
9.2 Distance-based approach
298(22)
Storing transactions as a graph
300(2)
Creating the k-nearest neighbors graph
302(7)
Identifying fraudulent transactions
309(9)
Advantages of the graph approach
318(2)
10 Social network analysis against fraud
320(37)
10.1 Social network analysis concepts
323(3)
10.2 Score-based methods
326(22)
Neighborhood metrics
330(6)
Centrality metrics
336(8)
Collective inference algorithms
344(4)
10.3 Cluster-based methods
348(6)
10.4 Advantages of graphs
354(3)
PART 4 TAMING TEXT WITH GRAPHS
357(74)
11 Graph-based natural language processing
359(30)
11.1 A basic approach: Store and access sequence of words
363(10)
Advantages of the graph approach
373(1)
11.2 NLP and graphs
373(16)
Advantages of the graph approach
387(2)
12 Knowledge graphs
389(42)
12.1 Knowledge graphs: Introduction
390(3)
12.2 Knowledge graph building: Entities
393(9)
12.3 Knowledge graph building: Relationships
402(7)
12.4 Semantic networks
409(6)
12.5 Unsupervised keyword extraction
415(13)
Keyword co-occurrence graph
423(2)
Clustering keywords and topic identification
425(3)
12.6 Advantages of the graph approach
428(3)
Appendix A Machine learning algorithms taxonomy 431(4)
Appendix B Neo4j 435(14)
Appendix C Graphs for processing patterns and workflows 449(9)
Appendix D Representing graphs 458(3)
Index 461
Alessandro Negro is a Chief Scientist at GraphAware. With extensive experience in software development, software architecture, and data management, he has been a speaker at many conferences, such as Java One, Oracle Open World, and Graph Connect. He holds a Ph.D. in Computer Science and has authored several publications on graph-based machine learning.