Implementing Analytics

Size: px
Start display at page:

Download "Implementing Analytics"

Transcription

1 Implementing Analytics A Blueprint for Design, Development, and Adoption Nauman Sheikh ELSEVIER AMSTERDAM BOSTON HEIDELBERG LONDON NEW YORK OXFORD PARIS SAN DIEGO SAN FRANCISCO SINGAPORE SYDNEY TOKYO Morgan Kaufmann is an imprint of Elsevier M<

2 Contents ACKNOWLEDGMENTS AUTHOR BIOGRAPHY INTRODUCTION Part 1 Concept xi xiü xv CHAPTER 1 Defining Analytics 3 The Hype 3 The Challenge of Definition 4 Definition 1: Business Value Perspective 5 Definition 2: Technical Implementation Perspective 6 Analytics Techniques 7 Algorithm versus Analytics Model 8 Forecasting 9 Descriptive Analytics 11 Predictive Analytics 13 Decision Optimization 18 Conclusion of Definition 20 CHAPTER 2 Information Continuum 21 Building Blocks of the Information Continuum 22 Theoretical Foundation in Data Sciences 23 Tools, Techniques, and Technology 24 Skilled Human Resources 24 Innovation and Need 25 Information Continuum Levels 25 Search and Lookup 26 Counts and Lists 27 Operational Reporting 28 Summary Reporting 29 Historical (Snapshot) Reporting 30 Metrics, KPIs, and Thresholds 31 Analytical Applications 33

3 Analytics Models 35 Decision Strategies 36 Monitoring and Tuning Governance 38 Summary 40 CHAPTER 3 Using Analytics 41 Healthcare 42 Emergency Room Visit 42 Patients with the Same Disease 43 Customer Relationship Management 44 Customer Segmentation 44 Propensity to Buy 45 Human Resource 46 Employee Attrition 46 Resume Matching 47 Consumer Risk 48 Borrower Default 49 Insurance 49 Probability of a Claim 50 Telecommunication 51 Call Usage Patterns 51 Higher Education 51 Admission and Acceptance 52 Manufacturing 52 Predicting Warranty Claims 53 Analyzing Warranty Claims 54 Energy and Utilities 54 The New Power Management Challenge 55 Fraud Detection 57 Benefits Fraud 57 Credit Card Fraud 57 Patterns of Problems 58 How Much Data 59 Performance or Derived Variables 59 Part 2 Design CHAPTER 4 Performance Variables and Model Development 63 Performance Variables 63 What are Performance Variables? 64 Designing Performance Variables 70 Working Example 73 Model Development 75

4 What is a Model? 75 Model and Characteristics in Predictive Modeling 75 Model and Characteristics in Descriptive Modeling 78 Model Validation and Tuning 79 Champion-Challenger: A Culture of Constant Innovation CHAPTER 5 Automated Decisions and Business Innovation 85 Automated Decisions 85 Decision Strategy 85 Business Rules in Business Operations 87 Decision Automation and Business Rules 88 Joint Business and Analytics Sessions for Decision Strategies 89 Examples of Decision Strategy 89 Decision Automation and Intelligent Systems 94 Learning versus Applying 94 Strategy Integration Methods 96 Strategy Evaluation 97 Retrospective Processing 97 Reprocessing 97 Champion-Challenger Strategies 98 Business Process Innovation 98 CHAPTER 6 Governance: Monitoring and Tuning of Analytics Solutions 101 Analytics and Automated Decisions 101 The Risk of Automated Decisions 102 Monitoring Layer 102 Audit and Control Framework 103 Organization and Process 103 Audit Datamart 104 Control Definition 106 Reporting and Action 108 Part 3 Implementation CHAPTER 7 Analytics Adoption Roadmap 113 Learning from Success of Data Warehousing 113 Lesson 1: Simplification 113 Lesson 2: Quick Results 114 Lesson 3: Evangelize 114 Lesson 4: Efficient Data Acquisition 115 Lesson 5: Holistic View 115

5 Lesson 6: Data Management 115 The Pilot 117 Business Problem 117 Management Attention and Champion 118 The Project 119 Results, Roadshow, and Case for Wider Adoption 125 CHAPTER 8 Requirements Gathering for Analytics Projects 129 Purpose of Requirements 129 Requirements: Historical Perspective 129 Calculations 130 Process Automation 132 Analytical and Reporting Systems 132 Analytics and Decision Strategy 133 Requirements Extraction 134 Problem Statement and Goal 135 Data Requirements 139 Model and Decision Strategy Requirements 142 Business Process Integration Requirements 144 CHAPTER 9 Analytics Implementation Methodology 147 Centralized versus Decentralized 148 Centralized Approach 148 Decentralized Approach 149 A Hybrid Approach 149 Building on the Data Warehouse 149 Methodology 151 Requirements 152 Analysis 153 Design 158 Implementation 164 Deployment 165 Execution and Monitoring 165 CHAPTER 10 Analytics Organization and Architecture 167 Organizational Structure 167 BICC Organization Chart 168 Roles and Responsibilities 170 Skills Summary 175 Technical Components in Analytics Solutions 176 Analytics Datamart 176

6 Contents CHAPTER 11 Big Data, Hadoop, and Cloud Computing 185 Big Data 185 Velocity 186 Variety 187 Volume 187 Big Data Implementation Challenge 188 Hadoop 189 Hadoop Technology Stack 189 Hadoop Solution Architecture 191 Hadoop as an Analytical Engine 193 Cloud Computing (For Analytics) 196 Disintegration in Cloud Computing 196 Analytics in Cloud Computing 197 CONCLUSION 199 REFERENCES 203 INDEX 207