Preface

15

Who Should Read this Book

15

Structure of the Book

16

Part I: Introduction

16

Part II: Building Machine Learning Applications

17

Part III: Use Cases and Roadmaps

18

Acknowledgments

18

Laboni Bhowmik

18

Avijit Dhar

18

Ranajay Mukherjee

18

1 Machine Learning and Intelligent Enterprise

23

1.1 What Is Machine Learning?

25

1.2 Transition from the Digital Era to the Intelligent Era

25

1.3 Intelligent Enterprise Use Cases

26

1.3.1 Intelligent Invoice Matching

26

1.3.2 Smart Recruiting

27

1.3.3 Service Ticket Intelligence

27

1.3.4 Product Similarity Matching

27

1.3.5 Brand Impact

28

1.3.6 Robotic Process Automation

28

1.3.7 Virtual Assistance

29

1.4 SAP’s Intelligent Enterprise Strategy

29

1.4.1 Intelligent Suite

30

1.4.2 Digital Platform

31

1.4.3 Intelligent Technologies

31

1.5 SAP’s Machine Learning Technologies and Applications

32

1.5.1 Embedded Machine Learning Applications

32

1.5.2 SAP Intelligent Robotic Process Automation

33

1.5.3 SAP Conversational Artificial Intelligence

34

1.5.4 SAP Data Intelligence

34

1.5.5 Embedded Machine Learning within SAP HANA

35

1.6 Summary

36

2 Machine Learning Fundamentals

37

2.1 Basic Probability Concepts

37

2.1.1 What is Probability?

38

2.1.2 Conditional Probability and Independence

47

2.1.3 Random Variables

49

2.1.4 Density and Distribution Function

54

2.1.5 Expectation

60

2.1.6 Variance and Standard Deviation

61

2.1.7 Covariance and Correlation

62

2.2 Basic Machine Learning Concepts

63

2.2.1 Different Types of Machine Learning Scenarios

63

2.2.2 Types of Learning

65

2.3 Machine Learning Algorithms

66

2.3.1 Linear Regression Analysis

66

2.3.2 Classification

85

2.3.3 Cluster Analysis

113

2.3.4 Other Machine Learning Techniques

125

2.4 Summary

137

3 Implementation Lifecycle

139

3.1 Understanding the Implementation Lifecycle

140

3.1.1 Cross Industry Standard Process for Data Mining

140

3.1.2 Define, Measure, Analyze, Improve, Control

141

3.1.3 Analytics Solutions Unified Method for Data Mining/Predictive Analytics

142

3.1.4 Key Components in a Machine Learning Project Lifecycle

142

3.2 Knowing the Business

143

3.3 Understanding and Exploring Data

144

3.3.1 Checking Data Quality

147

3.3.2 Checking Summary Statistics

151

3.3.3 Visualizing Data

153

3.4 Preparing Data

156

3.4.1 Data Conversion

157

3.4.2 Dimensionality Reduction

160

3.4.3 Variable Transformation

161

3.5 Developing the Model

163

3.6 Evaluating and Fine-Tuning Model

165

3.6.1 Model Overfitting

165

3.6.2 Model Performance

166

3.6.3 Model Fine-Tuning

168

3.6.4 Diagnostic Checks

168

3.6.5 Validation Set Approach

170

3.6.6 K-Fold Cross Validation

171

3.7 Deploying the Model

172

3.8 Summary

173

4 Machine Learning on SAP HANA

175

4.1 SAP HANA Machine Learning Components

175

4.1.1 SAP HANA Predictive Analysis Library

177

4.1.2 SAP HANA Automated Predictive Library

181

4.1.3 SAP HANA Extended Machine Learning Library

184

4.1.4 R Integration

188

4.2 Summary

204

5 Machine Learning with SAP Data Intelligence

205

5.1 Data Science Project Lifecycle

207

5.2 Managing the Data Science Project Lifecycle

209

5.3 SAP Data Intelligence

210

5.3.1 SAP Data Intelligence Versus SAP Data Hub

211

5.3.2 Architecture

212

5.3.3 Values of SAP Data Intelligence

215

5.4 Key Capabilities

216

5.4.1 Metadata and Governance

217

5.4.2 Machine Learning Scenario Manager

218

5.4.3 Machine Learning Data Manager

220

5.4.4 AutoML

221

5.4.5 Machine Learning Tracking

223

5.4.6 SAP Data Intelligence Modeler and Built-in Operators

227

5.5 Migrating to SAP Data Intelligence from SAP Data Hub

235

5.6 Summary

236

6 SAP HANA Predictive Analysis Library and R Integration

241

6.1 SAP HANA Predictive Analysis Library

241

6.1.1 Prerequisites and Installation

241

6.1.2 SAP HANA PAL Procedures

243

6.1.3 Calling SAP HANA PAL Procedures

250

6.1.4 Model Evaluation

260

6.2 R Integration

266

6.2.1 Prerequisites and Installation

266

6.2.2 R Packages

267

6.2.3 Calling R Functions from SAP HANA

268

6.2.4 Model Evaluation

275

6.3 Summary

278

7 Developing Applications with SAP HANA Predictive Analysis Library

279

7.1 Introduction to the Use Case

279

7.2 Building a Predictive Analytics Application Using SAP HANA PAL

280

7.2.1 Understanding the Business Problem and the Data

281

7.2.2 Data Preparation

291

7.2.3 Model Building

293

7.2.4 Model Validation and Refinement

300

7.2.5 Model Deployment

307

7.2.6 Interpretation and Displaying Results

311

7.3 Summary

315

8 SAP AI Business Services

317

8.1 Overview

318

8.2 Document Classification

319

8.2.1 Test Run of the Document Classification Service

322

8.3 Document Information Extraction

332

8.3.1 Test Run of the Document Information Extraction Service

335

8.4 Business Entity Recognition

339

8.5 Data Attribute Recommendation

341

8.5.1 Test Run of the Data Attribute Recommendation Service

344

8.6 Invoice Object Recommendation

347

8.7 SAP Service Ticket Intelligence

348

8.7.1 Test Run of the Service Ticket Intelligence Service

350

8.8 Summary

352

9 Building Scenarios Using Jupyter Notebook

353

9.1 Adding a Notebook

354

9.2 SAP Data Intelligence Python SDK

357

9.2.1 Setting Up and Leveraging the SAP Data Intelligence SDK

358

9.2.2 SAP Data Intelligence Python SDK

358

9.2.3 Popular Python Libraries

359

9.2.4 Various Library Options

360

9.3 Use Case

361

9.3.1 Setting Up the Scenario

363

9.3.2 Scripting

367

9.3.3 Data Plotting

371

9.3.4 Code Execution

372

9.4 Summary

374

10 Automated Machine Learning Data Science Automation

375

10.1 AutoML on SAP Data Intelligence

376

10.2 Features of AutoML

376

10.3 AutoML Step-by-Step

377

10.3.1 Configuring the Training Data Collection

378

10.3.2 Configuring the Test Data Collection

384

10.3.3 Setting Up AutoML

385

10.3.4 Testing AutoML

395

10.4 Summary

397

11 Conversational Artificial Intelligence

399

11.1 Introduction to SAP Conversational Artificial Intelligence

399

11.2 SAP Conversational AI

401

11.2.1 Natural Language Processing Engine

402

11.2.2 Intents

404

11.2.3 Expressions

405

11.2.4 Entities

406

11.2.5 Languages

409

11.2.6 Channels

412

11.3 Bot Building Techniques

412

11.3.1 Bot Building Process

413

11.3.2 Bot Connector

417

11.3.3 Bot Monitoring and Analytics

418

11.3.4 SAP Conversational AI APIs

419

11.4 Building a Chatbot Using SAP Conversational AI

421

11.4.1 Let’s Build a Bot!

423

11.4.2 Setting Up the Bot Project

424

11.4.3 Creating Intents

425

11.4.4 Setting Up Entities

427

11.4.5 Setting Up Skills

429

11.4.6 Enabling Voice

435

11.5 Summary

436

12 Integrating Machine Learning with the Internet of Things and Blockchain

439

12.1 Technology-Driven Transformation

441

12.2 Data—The Common Theme

442

12.3 Use Cases

445

12.3.1 Supply Chain

445

12.3.2 Healthcare

450

12.4 Summary

456

13 Industry Use Cases for Machine Learning Applications

457

13.1 Acceptance of Machine Learning across Different Industries

457

13.1.1 Assessing the Impact of Machine Learning

458

13.1.2 Challenges of Machine Learning Applications

460

13.2 Machine Learning Ecosystem

461

13.3 Identifying Industry Use Cases

464

13.3.1 Manufacturing Industry

465

13.3.2 Customer Service Industry

469

13.3.3 Healthcare Service Industry

473

13.3.4 Finance Industry

476

13.4 Summary

480

14 Conclusion and Roadmap

481

14.1 Recap

481

14.2 Best Practices

483

14.3 Roadmap

484

14.4 Summary

486